WIWWW“WWIHHHHIIHHIWH’Ml

   

THES!S

IVERSITY LIBRARIE

Ilillllllllnllll illlllllllll Hill

31293 017141247

ll“

 

 

 

 

This is to certify that the
thesis entitled
Comparison of Bone Growth in Stall- Versus

Pasture-Reared Horses

presented by

Kari Elaine Hoekstra

has been accepted towards fulfillment
of the requirements for

ILS. degree in Annual Sc1ence

 

 

5?; Jazz“

Major professor

Date ZNZ/j /?QX
/

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY

Mlchlgan State
Unlverslty

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

MTE DUE MTE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

COMPARISON OF BONE GROWTH IN STALL- VERSUS PASTURE-REARED
HORSES

By

Kari Elaine Hoekstra

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Animal Science

1998

ABSTRACT

COMPARISON OF BONE GROWTH IN STALL- VERSUS PASTURE-REARED
HORSES

By

Kari Elaine Hoekstra

Sixteen Arabian yearlings were randomly assigned to two experimental groups,
stalled and pastured, to investigate the efi‘ects of stalling versus pasture-rearing on bone
growth over a 140-day period Following an 84-d pre-training period, six horses from
each group were randomly chosen to complete a 56-d training period Serum osteocalcin,
Ca, P, 25-hydroxyvitamin D, and parathyroid hormone concentrations were determined
from blood samples taken every 14 d. Urinary deoxypyridinoline, pyridinoline, Ca, and P
concentrations and mineral content of the third metacarpal, as determined by radiographic
bone aluminum equivalencies (RBAE), were determined every 28 d from 24-hr urine
collections and radiographs of each horse’s left fi'ont leg, respectively. Lateral RBAE was
lower in the stalled horses at d 28 and remained lower throughout most of the project,
while pastured horses had increasing lateral RBAE. Horses kept in stalls had lower RBAE
of the mdial cortex at d 28. Medial RBAE tended to remain lower in stalled horses than
in pastured horses throughout most of the project. Serum osteocalcin concentrations were
lower and urinary deoxypyridinoline concentrations were higher in the stalled horses at d
14 and d 28, respectively, compared with the pastured horses, and subsequently returned
to baseline. Results suggest that housing yearling/two-yr-old horses in stalls may

negatively afl‘ect normal bone growth experienced by horses maintained on pasture.

Dedicated to my mother, father, and sister.

iii

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................... vi

LIST OF FIGURES ...................................................................................................... viii

INTRODUCTION ........................................................................................................... 1

CHAPTER 1

REVIEW OF LITERATURE .......................................................................................... 2
Bone Architecture ................................................................................................... 2
Bone Modifications ................................................................................................. 8
Skeletal Response to Exercise ............................................................................... ll
Skeletal Response to Limited Physical Activity ...................................................... 13
Measuring Skeletal Changes .................................................................................. 16

CHAPTER 2

MATERIALS AND METHODS ................................................................................... 25
Mamgement of Animals ........................................................................................ 25
Training of Animals ............................................................................................... 26
Sample Collection ................................................................................................. 27
Urine Collections .................................................................................................. 28
Elective Exercise Observations .............................................................................. 28
Sample Amlysis .................................................................................................... 29
Radiograplm .......................................................................................................... 30
Statistical Analysis ................................................................................................ 31

CHAPTER 3

RESULTS ..................................................................................................................... 33
Weight, Height, Third Metacarpal Circumference, and Body Condition Score ....... 33
Elective Exercise ................................................................................................... 34
Radiographic Bone Aluminum Equivalence ........................................................... 34
Serum Osteocalcin ................................................................................................ 36
Urinary Deoxypyridinoline and Pyridinoline ........................................................... 37
Serum Minerals, Vitamins, and Hormones ............................................................. 40
Urinary Minerals ................................................................................................... 40

CHAPTER 4

DISCUSSION ............................................................................................................... 43

CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS .......................................................... 60

iv

APPENDIX A ............................................................................................................... 64

Means Tables ........................................................................................................ 64
APPENDIX B ............................................................................................................... 67
Proc Mixed Tables ................................................................................................ 67
LITERATURE CITED .................................................................................................. 71

LIST OF TABLES

Table 1. Calculated analysis of total ration on an as-fed basis ......................................... 25
Table 2. Calculated daily elective strides per horse from observation ............................. 34
Table 3. Meamtable forlateralRBAE (mmAl) ............................................................ 35
Table 4. Means table for medial RBAE (mm Al) ........................................................... 35
Table 5. Means table for total RBAE (mm: Al) ............................................................. 37
Table 6. Means table for minary pyridinoline (nM/mM) ................................................ 39
Table 7. Means table for serum calcium (mg/d1) ............................................................ 41
Table 8. Means table for serum phosphorus (mg/d1) ...................................................... 41
Table 9. Means table for urinary calcium (g/d) .............................................................. 42
Table 10. Means table for urinary phosphorus (g/d) ...................................................... 42
Appendixka 1A. Meanstableforbodyweight (kg) .................................................. 64
Appendix Table 2A. Means table for height at withers (cm). ......................................... 64
Appendix Table 3A. Means table for third metacarpal circumference (cm) ..................... 64
Appendix Table 4A. Means table for body condition score. ........................................... 64
Appendix Table 5A. Means table for change in lateral RBAE (mm Al). ......................... 64
Appendix Table 6A. Means table for change in mdial RBAE (mm Al). ........................ 65
Appendix Table 7A. Means table for serum osteocalcin (ng/ml). ................................... 65
Appendix Table 8A. Means table for urinary deoxypyridinoline (nM/mM). .................... 65
Appendix Table 9A. Means table for urinary deoxypyridinoline (g/d). ........................... 65
Appendix Table 13. Proc mixed table for body weight (kg). ......................................... 67
Appendix Table 2B. Proc mixed table for height at withers (cm). .................................. 67
Appendix Table 3B. Proc mixed table for third metacarpal circumference (cm). ............ 67

Appendix Table 4B. Proc mixed table for body condition score. .................................... 67

Appendix Table 5B. Proc mixed table for lateral RBAE (mm Al). ................................. 67
Appendix Table 6B. Proc mixed table for medial RBAE (mm Al) .................................. 67
Appendix Table 7B. Proc mixed table for change in lateral RBAE (mm Al). .................. 68
Appendix Table 8B. Proc mixed table for change in medial RBAE (mm Al). ................. 68
Appendix Table 9B. Proc mixed table for total RBAE (mm2 Al) .................................... 68
Appendix Table 10B. Proc mixed table for serum osteocalcin (ng/ml). .......................... 68
Appendix Table 11B. Proc mixed table for urinary deoxypyridinoline (nM/mM). ........... 68
Appendix Table 12B. Proc mixed table for urinary deoxypyridinoline (g/d). .................. 68
Appendix Table 13B. Proc mixed table for urinary pyridinoline (nM/mM). .................... 69
Appendix Table 14B. Proc mixed table for serum calcium (mg/d1) ................................. 69
Appendix Table 15B. Proc mixed table for serum phosphorus (mg/d1) ........................... 69
Appendix Table 16B. Proc mixed table for urinary calcium (g/d). .................................. 69
Appendix Table 17B. Proc mixed table for urinary phosphorus (mg/d). ......................... 69

vii

LIST OF FIGURES

Figure 1. Change in Lateral RBAE (mm Al) versus day ofproject ................................. 35
Figure 2. Change in Medial RBAE (mm Al) versus day of project ................................. 36
Figure 3. Serum osteocalcin (ng/ml) versus day of project ............................................. 38
Figure 4. Urinary deoxypyridinoline (nM/mM) versus day of project ............................. 38
Figure 5. Urinary deoxypyridinoline (g/d) versus day of project ..................................... 39

viii

INTRODUCTTON

A major concern with young performance horses is the high incidence of skeletal
injury. The practice of transferring young, growing horses fi'om pasture to stalls prior to
yearling sales or commencement of training may predispose them to injury. Previous
researchlnsdemonstratedadecreaseinbonemineralcontent ofthethirdmetacarpalin
young horses soonaflertheonset ofrace training, aswellasachange inhousing fi'om
pastureto stalls(Nielsenetal., 1997). Thedecreaseinbonemineralcontentcouldresult
fiomincreasedordecreasedstrainrates onthebone associatedwithtrainingortheclmnge
in housing. Transferring young horses from pasture to stalls results in a slowdown in the
rate of bone formation due to a decrease in physical activity (Maenpaa et aL, 1988).
Studies ofother speciesdemonstrated similardecreasesinbone strengthinresponseto
confinement rearing (Knowles and Broom, 1990; Marchant and Broom, 1996). Thus,
yearling horses maintained on pasture with free access to exercise may have a skeletal
stuctumMisbeflerprepmedforthreasedbiomchmicalforcesMoccmdmmg
training and competition than horses housed in stalls with limited access to exercise.

The objectives of this study were to determine if bone development is negatively
afi‘ectedwhenyearlingsaretakenfiompastureto behousedinstallsandallowed limited
exercise, and to determine the consequential effects of the change in housing on bone

modeling/remodeling at the onset of training.

CHAPTER 1

REVIEW OF LITERATURE

Bone Architecture

All bones are composed of both cortical and cancellous bone (Marks and Popofl‘,
1988). Cortical bone nukes up the majority of the total skeletal rmss (Ice, 1988).
Characterized by its compactness, cortical bone forms the outer shell of all bones. In
contrast, cancellous bone, characterized by a fiamework of rods, plates, and arches,
individuallycalledtrabeculae, andahighbonemarrowcontent, isfoundprimarilyinthe
interior of bones. For example, the diaphysis, or central cylindrical shaft, of the third
metacarpaliscomposed primarilyofcorticalbone surroundingannrroweavity. The
epiphyses, locatedateitherendinlongbones,mdmetaphyses,themgionsconmctingthe
diaphysis and epiphyses, are mainly made up of trabecular bone surrounded by an outer
layer of cortical bone. The cavities of cancellous bone and the bone marrow in the
diaphysis are lined with the endosteum, a membranous layer of differentiated bone cells
(osteoblasts and osteoclasts). The periosteum on the outer surfaces of most bones are
lined with a layer of undifferentiated cells (mesenchymal stem cells). When stimulated,
thesecellscanberecruitedto increasebone growthand facilitate fi'acturerepair. The
periosteum itselfis covered by a thin layer of connective tissue.

Bone tissue exists in either a primary or a secondary form (Ice, 1988). Primary, or
immature, bone, also referred to as woven bone, is characterized by collagen fibers that
lack an ordered arrangement, osteocytes that are fewer in number and less uniformly

placed within the bone extracellular nutrix, and a lower mineral content, when compared

with secondary bone. Prinnry bone is present in the skeleton when developing bone tissue
initially forms, an injury occurs, or a skeletal abnormality is present. Secondary, or
lamellar, bone, with its patterned orientation of collagen fibers, evenly distributed
osteocytes, and high mineral content, replaces most prirmry bone. Both primary and
secondary bone can be either cortical or cancellous.

The osteon, orHaversiansystem, isthemainstructm'alunitofcorticalbone (Ice,
1988). Each osteon is comprised of a large, central vascuhr channel, or Haversian cam],
surrounded by circular sheets of collagen fibers, or concentric lamellae. Haversian canals
themselves contain blood vessels, nerves, and connective tissue. Volkmann’s canals, set
transversely to the longitudinal Haversian canals, form a network for osteocytes that
provides nutrients and allows intercellular signaling. Osteocytes are the main cells of
mature bone, located within the central canals, formed from osteoblasts that have
surrounded thermelves with mineral (Currie, 1988). Osteocytes can regulate mineral
homeostasis. Volkmann's canals support the osteocytes through connections to bone
surface cells, bone rmrrow, and other Haversian canals (Jee, 1988). Cancellous bone does
not contain Haversian systems. Instead, it contains trabecular packets within its
framework, or trabeculae, that are fimctionally analogous to the osteon. Throughout the
lamellae of both cortical and cancellous bone are srmll cavities, or lacrmae, occupied by
osteocytes and connected by thin tubular channels, or canaliculi (Marks and Popofl‘,
1988). The canaliculi contain the cytoplasmic processes of the osteocytes, which allow
communication with other osteocytes through gap junctions and provide nutrients through
connections to the bone surface. In addition to osteocytes, other cells present in bone are

osteoblasts, which are found along active bone surfaces, and osteoclasts, which usually

reside in or near Howship’s lacunae, the resorption pits located on the bone surface (Ice,
1988). Osteoblasts are bone-forming cells that synthesize and secrete unmineralized bone
matrix (osteoid), participate in calcification of bone, and regulate mineral homeostasis
throughthe fluxofcalciumandphosphate inandout ofbone. Osteoclastsaregiant
multinucleated cells, originating from bone rmrrow, whose primary function is the
resorption of bone, either internally or on the srn'face of bone (Jaworski, 1984; Weryha
and Leclere, 1995). Also present in bone are bone-lining cells. These thin, elongated cells
are found directly apposed to inactive bone surfaces (Marks and Popofl‘, 1988). Bone-
finhgceflsfimdionprhnafilyasanionbmfierseparatmgmtersfifialfluidfiommebom
fluids ofthe lacunar-canalicular system. Bone-lining cells are postulated to be involved in
sensing the nugnitude, distribution, and rate of mechanical strain placed on bone during
loadingandnansmntmgthismfornntionmtheacfiveboneceflsstnnuhfingthe
appropriate bone formation or resorption response (Ice, 1988). Osteocytes lave also been
hypothesized to be responsible for this transmittance (Lanyon, 1987), due to their wide
distribution throughout the organic matrix and their ability to communicate with each
other and other bone surface cells through canalicular processes (Ice, 1988).

The process of bone degradation by osteoclastic cells involves the release of
calcium and phosphate ions into the extracellular fluid, followed by proteolytic
degradation of the organic matrix of bone (Marks and Popofl‘, 1988). The surface of an
active osteoclast, characterized by its membrane infoldings, is called the ruflled border
(Ice, 1988). A clear zone surrounds the ruflled border and provides a tight seal for
maintemnce of an appropriate microenvironment for bone resorption through its adhesion

to the bone surface. Acid hydrolases stored in Golgi complexes, located in the cytoplasm

ofthe osteoclast cell, aretransported to therufifledborderregion, wherethemembraneof
meprhnarylysosomemseswimmmflledbordamdmhasestheacidhydrohsesmto
theexnaceflmmspacecreatedbetweentheboneandtheceflbytheckmmmmksmd
Popofl‘, 1988). This creates an acid nricroenvironment where the bone can be degraded by
acidic proteases. The breakdown products of the bone are taken up in digestive vacuoles
and secondary lysosomes, finther degraded, and released into vascular spaces nearby.
Once minerals are released, the organic nutrix is recorbed (Jee, 1988). Cells of the
osteoblast lineage are believed to play a role in the initiation of osteoclastic resorption by
their destruction of the osteoid tissue lining the bone nntr'ix (Weryha and Leclere, 1995).
Bone formation, on the other hand, involves matrix formation, followed by mineralization.
The synthesis of organic matrix by active osteoblasts occurs at the common boundary
between osteoblastic cells and osteoid that is already present (Jee, 1988). Mineralization
subsequemlyoccmsmmemterfacebaweenostwidandthennstrecemlymmaahzed
bone. The lagtimebetweenmatrixformationandmineralizationintheadultis
approximatelle days. During this time the matrix isundergoing a series ofsteps in
preparation for mineralization. The process of mineralization hsts several months and is
divided into two distinct phases (Jee, 1988). The primary phase occurs over several days
and results in approximately 70 percent of total mineralimtion. This pinse is thought to
be regulated by osteoblasts on the osteoid surface and osteocytes within the lacume of
osteoid. The secondary phase lasts several months and provides most of the remaining 30
percentofmineralization. Theavailabilityofmineraltothematrixandthechemical
composition ofthe fluid surrounding the matrix may be the governing elements ofthe

secondary phase of mineralization. The mechanism for the activation of mineralization can

be bone resorption or can occur independently of any local resorption (Jaworski, 1984).
Various interleukins, prostaglandins, growth factors, and hormones have been implicated
in the regulation of osteoclast and osteoblast cell activity (Weryha and Leclere, 1995).
Intheyoung, growinganirnal, cartilaginous growthplates, alsoreferredtoas
epiphyseal-metaphyseal complexes, separate the epiphyses and metaphyses of the long
bones. Proliferation, and hence elongation, of the long bone occurs at this site (Orrrie,
1988). The growth plate itself is made up of five zones, referred to as the resting,
proliferative, mturation, hypertrophic, and calcification zones (Ice, 1988). Each zone
provides a distinct and important frmction in the elongation of long bones. The resting
zone, closest to the epiphysis, firnctions as a source of progenitor cells, which have the
capacity to undergo mitosis and difl‘erentiation into chondrocytes (Currie, 1988). These
difi‘erentiated cells, assembled into longitudinal columns, comprise the proliferative zone.
Within this zone, difl‘erentiated cells actively undergo mitosis and synthesis and secretion
ofcartilagematrix. Thesynthesisofmatrixandthepreparationofthematrixfor
calcification occurs in the zone of cell maturation (Jee, 1988). In the hypertrophic zone,
cells are characterized by their enlargement and intracellular glycogen and calcirnn stores,
the latterofwhichisactivelyremoved fromthe cellsandused inthemineralimtionofthe
cartilage nutrix (Cm'rie, 1988). This initial step in the conversion of cartilage to bone
occurs in the calcification zone, located closest to the metaphysis. The next step in the
conversion involves blood vessels and osteoblasts entering the mineralized cartilage.
Osteoblasts secrete an organic matrix (osteoid) containing type I collagen, noncollagenous
proteins, phosphoproteins, and lipids, which is subsequently mineralized by the deposition

of inorganic salts, primarily calcium phosphate, to form bone (Currie, 1988; Marks and

Popofl‘, 1988). In essence, bone elongation occurs by the growth plate growing away
fi'om the diaphysis (Currie, 1988). Chondrocytes eventually die due to a lack of available
nutrients resulting fi'om matrix calcification, and any remaining cartilage is removed by
osteoclastic cells. Eventually this primary, or immature, bone, also referred to as the
spongiosa, is removed and replaced with secondary, or mature, bone at the region ofthe
metaphysis of the long bone. Growth at the epiphyseal-metaphyseal complex continues
untilcells ofthe resting zone canno longerundergo mitosisandchondrocytesareno
longer available to make up the proliferative zone. Upon cessation of growth, the band of
cartilage at the growth plate is completely replaced by cancellous bone and the adjacent
epiphysis and metaphysis fuse together. This process is referred to as closure of the
epiphyses. In the horse, epiphysial closme of the long bones will occur between 9 and 30
mo of age, depending on the specific bone (Evans et al., 1990).

While elongation occurs at the growth plate, the diaphysis of the long bone
develops circumferentially by a combination of bone deposition and degradation (Currie,
1988). Bone formation occurs on the bone surface, immediately internal to the
periosteum, while resorptiontakes place onthe inner surface ofthe shafiwall. Internal
resorptionwillresult inthe creationofacavityforbone marrow. Initially, whilethe
diameteroftheshafiexpandstheshaflwaflthinsasaresuhofmguhtedratesof
formation and degradation during growth. Ultimately, the wall thickens as bone formation
onthe surface ofthe bone accelerates, thus increasingthemechanicalstrengthofthe bone.
Theflaringoftheendsofalong bone during growthoccursbyformationofbone onthe
periosteal surface and osteoclastic resorption on the endosteal surface of the diaphyseal

shalt, while, at the same time, bone formation on the endosteal surface and periosteal

resorption are occurring in the metaphyseal region (Ice, 1988). This results in a long bone

with a narrow shall at thejunction ofthe metaphysis and flaring epiphysis, greater
diaphyseal diameter, and an enlarged marrow cavity.

Bone Modifications

The skeletal architecture functions prirmrily to protect the body’s soft tissues, such
astheskullandtheribs,andisdeterminedgenetically(Lanyon, 1987). Thegeneralshape
and composition of load-bearing bones, responsible primarily for the locomotion of the
skeleton, is also determined genetically. In the absence of functional loading, these bones
will develop into their recognizable form. However, only through an adaptive response to
physical loading will the structural features that enable these bones to withstand repetitive
loading without damage develop. These structural features include girth, cross-sectional
shape, cortical thickness, longitudinal curvature, and bone mineral content.

During bone growth, bone modeling and remodeling are occurring simultaneously.
The modeling mechanism, most obvious during long bone growth in a young animal,
causes modifications in the shape and size of the bone (Jee, 1988). In essence, modeling
determines the amount and form of bone in the adult animal’s body. Modeling occurs by
the apposition or resorption of bone on the bone surface. Examples of modeling include
drift, or shifting, ofthemidshafi, flaringoftheendsoflongbones, and increasingbone
volume during grth (Jaworksi, 1984). In contrast, the replacement of primary, or
immature, bone by secondary, or mature, bone, as well as the replacement of old or
damaged bone, is referred to as remodeling (Jee, 1988). In horses, approximately 50% of

thereplacementofimmaturebonewithmatureboneoccursby3yrofage,andthe

remodeling mechanism then continues throughout life, as secondary bone is continuously
being destroyed and replaced (Riggs and Evans, 1990). In remodeling, bone degradation
and formation occur at the same skeletal location, whereas either formation or resorption
occm in the modeling process, but at different surfaces and rates (Jee, 1988; Burr et al.,
1989). In essence, no change in the shape or gross amount ofbone present occurs with
the remodeling process because of the coupling of bone resorption and formation, whereas
anincreaseordecreaseinthetotalamount ofbone orachangeinshapeorlocationof
bone will occur with modeling, based on the independent activation of resorption or
formation (Jaworski, 1984).

Inadditionto sitesofimnntureandold bonereplacementtheareasofbonetlnt
endurethemost stressduring loading, especiallytendonattachment sites, andsustain
maximal loads, are at points of remodeling (Norwood, 1978). Three nnjor phases make
up the remodeling cycle. Osteoclastic resorption of primary, damaged, or old bone occurs
first, followed by a short reversal phase, during which the bone surfirce is prepared for
formation, and finally osteoblastic deposition of secondary or new bone. Layers of bone
may be removed and then replaced hour the surface ofthe bone, or osteoclasts may tunnel
their way through the inner cortex of bone, leaving behind a canal that is subsequently
filled by osteoblastic cells, producing a secondary Haversian system (Riggs and Evans,
1990). Burr et al. (1989) includes an activation phase, prior to resorption, at which time
theboneisstirnulatedto changeinresponseto mechanicalstressorstrainbythe
activation of preosteoclastic cells. The adaptive response of remodeling occurs by
engagement of new osteoblasts and/or osteoclasts, rather than increasing the activity of

alreadyestablishedcells. Thespecificmechanismthatlinksthemechanicalloading

stimulus to the activation of the bone cells responsible for the remodeling process is
unknown (Lanyon, 1989). The functional unit of the remodeling mechanimr is referred to
as the “bone remodeling unit” or “bone multicellular unit” (Jee, 1988). This unit includes
the group of cells responsible for a quantum of bone resorbed and replaced at a particular
site, as well as the quantum of bone involved.

Whentheboneisloaded, itdeformsduetothestrain.1f,whentheboneis
unloaded,theshapeoftheboneretumsto itsofiginalfomtheboneissaidto havebeen
loaded within its elastic region (Lanyon, 1989). Little modeling or remodeling occurs
withinthisregion. Whenthebonecanno longerresistthechangeinshapeandcontinues
to deform, in response to an increase in loading magnitude, it has reached a point beyond
the elastic region. Thisisreferred to asenteringtheyieldregionandoccurswhenthe
level of physical activity is increased to a point beyond what the skeletal structure is
accustomed. Continual loading within this region without allowing the skeleton to adjust
to the new level of activity through remodeling will eventually result in damage to the
bone. Norwood (197 8) has proposed that one remodeling cycle takes approximately 4 mo
to complete in the horse. The resorption phase is said to last approximately 1 mo, the
reversalphase about 1 wk, andthebone depositionphase approximately3 mo. Itis
during the period of time between the start of resorption and the completion of bone
deposition in the reparative remodeling mechanism that the bone is susceptible to injury
due to its weakness and porosity. The increase in porosity that occurs is related to the
number of Haversian systems that are being actively remodelled at the same time. The
more Haversian systems simultaneously undergoing remodelling, the greater the porosity

ofthebone,resultinginthestrainsthatareappliedtotheboneatthistimetobeincreased

10

(Nunamaker, 1986). If adequate time is not given for the bone to heal and modify itself
before additional stress is placed upon it, the risk of injury increases (Lanyon, 1984).
Damage can also result fi'om a single massive load applied to the skeletal system or fiom

fatigue faihrre due to repetitive loading (Lanyon, 1987).

Skeletal Response to Exercise

Wolfi’s Law states that as the biomechanical load on a bone changes, the interml
structureofthe bonewillmodifytoaccommodatethenewstressplaceduponit
(Norwood, 1978). While the general shape of load-bearing bones is genetically
determined, nrodeling and remodeling ensure that the structure of the load-bearing bones
issuflicient instrengthandrigidityto withstandrepetitiveloadingwithout damage
(Lanyon, 1987). In young, growing racehorses or other performance horses, the
adaptationofthe skeletalstructureisaresponsetothemechanicalloadsexperiencedasa
result of training. This adaptive response serves to develop optimal functional
characteristics of bone, as well as tendon and muscle, for the physical activity most
commonly encountered, which is the primary objective of training (Lanyon, 1989). Bone
cells responsrble for the adaptive response directly or indirectly recognize and
appropriately match the skeletal structure to the functional load applied during training
(Lanyon, 1987). As loading increases beyond the appropriate strain environment
previouslyutilized, aswhenthenext stageintrainingisintroduced, bone formation
predominantly occurs, decreasing the strain placed on the load-bearing bones by increasing
the amount of bone tissue present. Unless the magnitude, rate, or distribution of the strain

applied during training changes thereafter, finther modifications to the skeletal

ll

architecture are unnecessaryand subsequent loading ofthe sametypewillnot elicita
greater adaptive response (Lanyon, 1989). The skeletal structure is said to have reached
an optimal strain environment for that physical activity (Lanyon, 1984).

Woo et aL (1981) conducted a study to determine the effects ofa prolonged,
moderate intensity exercise program on the composition, mechanical properties, and
structural properties ofthe femur ofimmature swine. The study included both a non-
exercised control group and an exercised group that underwent 12 mo of conditioning on
a treadmill at an exercise level consistent with 65 to 80% of maxirmrrn heart rate. The
mechanical properties, bone density, and biochemical contents of the femurs in the non-
exercised and exercised groups were not difl‘erent after completion of the study.
However, cortical thickness, cross-sectional area, total volume, ash content, and calcium
content were significantly higher in the exercised group. Thus, the skeletal difi‘erences
betweenthe groups were attainedbyanincrease inbone quantity, withno afl‘ectonbone
quality. In response to mechanical loading, an adaptive mechanism was stimulated to
remodel the cortical bone, increasing the amount of bone tissue present and thereby
decreasing the strain encountered during exercise.

A study by McCarthy and Jefl‘cott (1992) examining the effects of treadmill
exercise on the equine third metacarpal bone showed similar results to that ofWoo and
colleagues (1981). Investigators in this study compared the structural properties and
cellular activity of cortical bone in the third metacarpal of two groups of yormg horses.
One group was kept relatively inactive through restricted housing and the other group was
subjected to an intense treadmill exercise program at near mximal speeds. Results

indicated that the cortical bone of the exercised horses experienced little bone remodeling,

12

but extensive bone formation occurred on the dorsal cortex, resulting in an increase in
thicknessofthis cortex. Thissuggeststhatthestrainsenduredbythethirdmetacarpal
during treadmill exercise were below that which would cause microdamage to the bone
and initiate bone remodeling, but were high enough to stimulate bone formation in an
attempt to adapt to the new functional load placed on the bone. Increasing the bone
mineral content of the dorsal cortex of the third metacarpal improved the bone’s ability to
withstand loading on that particular cortex.

Buckingham and Jefi‘cott (1991) compared the efl‘ects ofa long term submaximal
exercise program on the bone mass of a group of yearling Standardbreds to that of a non-
exercised control group. Bone mineral density increased in the exercised group, but
decreased in the non-exercised group, although the changes were small and lacked
significance. Even so, results indicated a possible trend toward increasing bone strength
anddensityasaresult ofexercise. What maybe more apparent fromtheresults ofthis
studyisthatwhile lowintensityexercisedidnotresultinahighdegree ofbone

hypertrophy, it was at least not detrimental.

Skeletal Response to Limited Physical Activity
An appropriate skeletal structure can only be maintained if mechanical loading is
performed on a regular basis (Lanyon, 1989). If loading falls beneath that normally
sustained by the skeletal enviromnent, bone formation will slow or cease and bone
resorption will predominantly occur until the skeletal structure and functional load applied
match (Rubin, 1984). Transferring young horses fi'om pasture to stalls during the winter

months has been shown to result in decreased serum osteocalcin concentrations (indicative

l3

of osteoblastic activity), indicating a slowdown in the rate of bone formation associated
with decreased physical activity (Maenpaa et al., 1988). Immobilization of young,
growing rats by sciatic denervation was shown to increase bone resorption and
dramatically decrease bone formation over 6 wk, resulting in a lower bone mineral content
(Yeh et al., 1993). Kannus et al. (1996) demonstrated a marked decrease in osteocalcin
irnmunoreactivity in the rat patella after 3 wk of irmnobilization, indicating reduced bone
formation. These examples support the existence of an adaptive response between
mechanical loading and the skeletal structure. If the fimctional strain on the load-bearing
limbs is suddenly reduced, the skeletal architecture responds by lowering bone mass to a
level appropriate for the new strain (Lanyon, 1984).

In the extreme case of complete skeletal disuse, Rubin and Lanyon (1984)
determined that the predictable response is primarily bone resorption and decreased bone
mineralcontent. Removal ofthe natural strainloading ofanintactroosterulnaby
functional isolation resulted in a 12% decrease in bone mineral content over 6 wk. An
insignificant amount of new bone formation did occur, due to the coupling of resorption
and formation in the remodeling mechanism. A similar study in humans showed a 10.4%
decrease inbone mineralofthe calcaneusandatotalbodybonemineralloss of1.4% in
healthy adult men voluntarily subjected to 17 wk ofbed rest (LeBlanc et al., 1990). In the
study by Rubin and Lanyon (1984), only four strain cycles per day, applied extermlly,
were necessary to prevent bone resorption and maintain a functional level of bone mass.
This indicates tlmt a firnctional strain environment for the load-bearing bones can only be
maintained if loading occurs on a continuous basis.

Placing yearling horses in stalls in preparation for yearling sales or the

14

commencement of training, without any exercise more rigorous than hand-walking, is
common in the horse industry. Few, if any, studies have examined the skeletal effects of
stall-rearing yearlings. In addition to the study by Maenpaa et al. (1988) looking at the
transferoffoals fiompastureto stallsanditsefi‘ects onbone formation, concemaboutthe
effects of stalling on bone growth stems fi'om various research studies discussing the
efi‘ects of confinement rearing on other livestock species. Laying hens housed in battery
cages were formd to have only 54% ofthe humeri strengthofbirdsthat were kept ina
perchery (Knowles and Broom, 1990). Fleming et al. (1994) showed that the breaking
strength and radiographic density of the humeri fi'om battery caged laying hens were 40 to
50% lower than values obtained for birds housed in two difi‘erent perchery systems. A
similardecrease inhumerusbreakingstrengthof45%wasdemonstratedincagedbirdsin
a study by Norgaard-Nielsen (1990). Similarly, sows housed in stalls were found to have
only two-thirds of the humeri and femur breaking strength of group-housed sows
(Marchant and Broom, 1996).

Research suggests that limiting a young, growing horse's access to physical activity
of such a magnitude as to signal the bone to remodel and become stronger may have
negative consequences. These consequences may include the development of a skeletal
systemthat isofinferiorstrengthandthusunpreparedto handletrainingandcompetition.
Young horses should enter training with a skeletal system that is optirmlly prepared for
the strain that mechanical loading will place upon it. Frost (1987) states that cortical bone
depositswillbeincreasedbythemodelingmechanismiffirnctionalloading ofboneis
vigorously increased in the juvenile skeleton. Additionally, future deposits will be retarded

if disuse of bone occurs. If the modeling mechanism primarily occurs in the young,

15

growinganirmLchangingthearchitectmeofthe bonewhiletheanimalisyoungandstill
capable of altering bone structure would be advantageous. This may contribute to the
prevention of career-threatening, or even life-threatening, injuries. However, if the
skeletal systems of young performance horses are being ill-prepared for training and
competition due to stalling, they may be predisposed to injury, which could result in a
shortened, or even terminated, competitive career. Determining if bone development is
negatively affected when young, growing horses are taken ficm pasture to be housed in
stalls and allowed limited exercise would be financially and emotionally advantageous for

the horse industry.

Measuring Skeletal Changes

Current research is being conducted to assess the capabilities ofbiochemical
markers of bone turnover in horses as a supplement to other noninvasive methods of
assessing bone mineral content, including radiographic photometry, photon
absorptiometry, and ultrasonography. Development of such biochemical markers would
provide a complimentary, and perhaps more sensitive, means of identifying equine athletes
at risk of injury. Biochemical markers investigated in this study are bone matrix
components released into the blood or urine during the formation of new bone by
osteoblasts or the resorption of old or damaged bone by osteoclasts, respectively (Price et
al., 1995; Delrnas, 1993). In young, healthy individuals, bone formation exceeds bone
resorption. In physically active adults, bone formation and degradation are tightly
regulated to maintain bone rmss. Development of bone marker assays may aid in

detecting situations when the skeletal system is more actively engaged in bone resorption

16

tlmn production and thus the risk of skeletal injury is higher. Assays for biochemical
indicators ofboneturnoverinserumandurinealreadyexist inhurnanmedicine forthe
clinical investigation of osteoporosis and other metabolic bone diseases (Delmas, 1993).
Theseassaysarenowbeing utilized forequineresearch. Inthisstudy,thebiochemical
indicators of bone formation and bone resorption analyzed were serum osteocalcin and
urinary pyridinoline and deoxypyridinoline, respectively.

Osteocalcin, or bone gla protein (BGP), is one of the most abundant
noncollagenous proteins of mammalian bone (Gomez et al., 1994). Osteocalcin contains
three amino acid residues of the vitamin K-dependent gamma-carboxyglutamic acid and an
associated disulfide bond (Price, 1982). The 49 amino acid protein has weak calcium
(Ca2+) binding properties, but a strong aflinity for hydroxyapatite (Gomez et al., 1994;
Price, 1982). The role of gamma-carboxyglutamic acid residues in osteocalcin is to fully
enable osteocalcin to bind to hydroxyapatite. Decarboxylation of gamma-carboxyglutamic
acid to glutanmte greatly weakens osteocalcin's Ca2+ binding properties and lowers its
aflinity for hydroxyapatite, although the non-gamma carboxylated protein will continue to
bind hydroxyapatite, albeit weakly (Price, 1982). Osteocalcin does not appear in
mineralized tissues until the initial mineral phase has matured into hydroxyapatite,
indicating the importance of hydroxyapatite binding to the accumulation of the protein in
bone (Price, 1982). The kidney is the main route of excretion for circulating osteocalcin,
following proteolysis (Gomez et al., 1994).

Osteocalcin has been found in osteoblasts, osteocytes, and dentin (Gomez et al.,
1994). However, certain characteristics of osteocalcin suggest that osteoblasts are the

cells within bone that are responsible for osteocalcin production. First, osteocalcin

l7

contains a 4-hydroxyproline residue in its structure, indicating the presence of prolyl
hydroxylase (Price et al., 1976). The prolyl hydroxylase enzyme has been used as a
marker for distinguishing osteoblasts fi'om osteoclasts in culture. Second, osteocalcin has
been shown to be synthesized exclusively by clonal osteosarcoma cells that exhibit features
of the osteoblast phenotype, such as high alkaline phosphatase activity and a high degree
of responsiveness to PT'H (Nishimoto and Price, 1980). Thus, measurements of serum
osteocalcin appear to directly reflect the activity of osteoblastic cells (Price, 1982).

Osteocalcin is synthesized by bone cells at new bone mineralization sites. A
Man of newly synthesized osteocalcin that fails to bind to the organic matrix of bone
dining bone formation is released intact fiom these mineralization sites into blood, where
it becomes a measurable marker of osteoblastic activity (Gomez et al., 1994; Price, 1982).
Osteocalcin is not released fi'om the bone matrix at any other time except new bone
formation (Price, 1982). The specific function of osteocalcin in bone metabolism is
unknown, although it is likely that osteocalcin regulates mineral formation. Its synthesis
by osteoblastic cells supports a role in bone formation, however the hormonal regulation
of osteocalcin synthesis complicates this idea. Osteocalcin synthesis is stimulated by 1,25
dihydroxyvitamin D3 (1,25-(OH)1D3), the active form of vitamin D, although osteocalcin
synthesis will occur in the absence of vitamin D. However, 1,25-(OH)2D3 also inhibits
collagen synthesis, the basic unit of bone formation (Price, 1982). The fimction of 1,25-
(OH)2D3 is to adjust bone metabolism during periods of stress involving inadequate
mineral levels from dietary intake through its stimulation of osteocalcin synthesis, implying
that it helps regulate mineral homostasis during bone resorption.

Brown et al. (1984) determined that serum osteocalcin correlated positively with

18

markers of osteoblastic activity, including relative osteoid volume, relative osteoid
surfaces, tetracycline labeled surfilces, and bone formation rate. Interestingly, serum
osteocalcin and bone resorption surfaces were not positively correlated. As an increase in
resorptionsurfacesisindicative ofanincrease inosteoclast cellactivity, the lack ofa
positive correlation with osteocalcin supports the idea that osteoblasts, and not
osteoclasts, synthesize osteocalcin. Carter et al. (1996) reported an inverse correlation
between osteocalcin and growth rate (P < 0.01), bone strength (P < 0.01), metacarpal ash
(P < 0.10), femur ash (P < 0.01), and femur ash weight (P < 0.01). Interpretation ofthese
results by the investigators indicated the high specificity of osteocalcin as a marker of bone
mineralization and/or turnover.

Osteocalcin’s direct correlation with bone metabolism and the relative ease of
radioirnmunoassay measurement have led to its attractiveness as a means of evaluating
bone disorders and the effects of physical activity on bone formation (Price et al., 1980).
When compared to two other commonly used biochemical rmrkers, phsma alkaline
phosphatase activity and urinary hydroxyproline, osteocalcin has an advantage in that it is
a specific bone protein, whereas alkaline phosphatase and hydroxyproline both have
several tissues of origin. In human medicine, serum osteocalcin can monitor changes in
the rate of bone formation conrrnon in patients with metabolic bone disease (Brown et al.,
1984; Delrnas, 1993). Osteocalcin is increased in patients with bone diseases
characterized by increased bone resorption and increased bone formation, including
Paget’s disease, bone metastases, primary hyperparathyroidism, renal osteodystrophy, and
osteopenia (Price et al., 1980). Brown et al. (1984) determined that serum osteocalcin

reflects bone fornntion, but not bone resorption, in patients with postmenopausal

l9

osteoporosis.

Lepage et al. (1990) measured serum osteocalcin concentrations in 50 clinically
normal female Standardbred horses of varying ages to determine difi‘erences in serum
levels with age. Mean osteocalcin concentrations for animals less than 1 yr of age,
between 1.5 and 2.5 yr, and older than 3.5 yr of age were 47.3, 35.7, and 6.7 ng/mL,
respectively. These results indicated that serum osteocalcin concentrations decrease with
age in fennle horses, suggesting a significant slowdown in the rate of bone formation in
adult horses compared to young horses. A higher degree ofvariation occurred in the 1.5
to 2.5 yr age group, which was attributed to differences in adaptation of bone turnover to
exercise, since 18 mo of age corresponded to the commencement of training

Another study examinedthe efl’ects oftransferring foals fi'ompastureto stallsfor
the winter months on biochemical markers of bone formation (Maenpaa et al., 1988). A
23.4% decrease in serum osteocalcin concentrations occurred within 1 mo of when the
foals were stalled. Investigators interpreted this decrease to indicate a slowdown in the
rate of bone formation in response to decreased physical activity associated with stalling.

Urinary pyridinoline and deoxypyridinoline, pyridinium cross-links of collagen, are
biochemical indicators of bone resorption or collagen degradation (Robins et al., 1994).
Pyridinolineispresentinboneandcartilagenntrix,aswellasinotherconnectivetissues,
while deoxypyridinoline is present in bone collagen of the organic mtrix (Delnms, 1993;
Robins et al., 1994). The function of pyridinoline and deoxypyridinoline is to provide
structural rigidity and strength for bone collagen. These crosslinks provide stability due to
their location between adjacent collagen fibers within the extracellular matrix (Gomez et

al., 1996). When collagen is degraded by various proteases, crosslinks are released into

20

circulation and ultirmtely are excreted in urine. Studies have shown that the
concentrations of pyridinoline and deoxypyridinoline found in urine are essentially derived
fi'ombone, based onasirnilarmolarratio ofpyridinoline to deoxypyridinoline inboneand
urine, therefore explaining their use as indicators of bone resorption (Gomez et al., 1996;
Robins et al., 1994).

Urinary excretion of pyrldmolrne and deoxypyridinoline has been shown to be
lmafl‘ected by renal dysfirnction (Robins et al., 1994). However, arthritic conditions in
humans have been shown to significantly increase levels of pyridinoline from somces other
than bone, suggesting that deoxypyridinoline may be a more specific and reliable marker
for bone resorption. Once released into the circulation, pyridinoline and deoxypyridinoline
cannot be reutilized in collagen production because both are products of a
posttranslational modification of collagen molecules that have already been secreted and
become part of the extracellular matrix (Dehnas, 1993). Pyridinoline and
deoxypyridinoline are not metabolized prior to excretion by the kidneys and are unafl‘ected
by diet, adding further to their reliability as indicators of bone degradation. Approximately
40% ofthe pyridiniumcrosslinksareexcreted inurine inafiee formand 60% ina
peptide-bound form (Delmas, 1993).

High-performance liquid chromatography (HPLC) has been used to measure total
amounts of crosslinks, while direct enzyme immunoassay (ELISA) methods have been
developed to measure free pyridinoline and deoxypyridinoline, with no significant
interaction with the peptide-bound form of the crosslinks (Robins et al., 1994). Studies
have shown that results from immunoassay methods are highly correlated with results

from I-IPLC assays measuring total crosslinks (Gomez et al., 1996; Robins et al., 1994).

21

Measures of pyridinium crosslinks have been found to be useful in human medicine for
assessingmetabolie bone diseasesandriskofdisease, aswellasmonitoringtherapy
(Robins et al, 1994). Concentrations of fiee pyridinoline and deoxypyridinoline in the
urinearehigherinchildrenthanadults, dueto greateractivityofthebonemodelingand
remodeling mechanisms during growth (Gomez et al., 1996; Robins et al., 1994). These
findings suggestthat measurements ofthe pyridiniumcrosslinksmaybe usefirl monitors of
growth and the presence of growth abnormlities (Robins et al., 1994).

A study comparing the effects of estrogen alone and estrogen and androgen
together on biochemical markers of bone formation and reSorption in postmenopausal
women included pyridinoline and deoxypyridinoline among the markers measured (Raisz
et al., 1996). Owing to the fact that estrogen deficiency is known to play a role in the
changesthat occurinbone metabolismaftermenopause, andandrogens are suspectedto
beinvolvedaswell, bothreplacementtherapiesusedinthisstudywereexpectedtoafl‘ect
both bone formtion and bone resorption. Results indicated that both therapies showed a
similar decrease in urinary excretion of deoxypyridinoline and pyridinoline, interpreted as
an indication of a slowdown in the rate of bone resorption.

The applicability of these biochemical mkers of bone formation and resorption to
the horse have exciting possrhilities, in terms of determining normal bone growth patterns
and identifying abnormal bone breakdown. The impact tint these markers may have on
reducing the potential for career- and life-threatening injuries in horse racing and other
equine performance activities is eneomaging.

Supplemental to the use of biochemical markers of bone metabolism are various

noninvasive methods of evaluating bone. Radiographs can be used for both qualitative

22

and quantitative evaluation of bone. Meakim et al. (1981) has developed a method to
determine radiographic bone aluminum equivalence (RBAE), a measurement of bone
mineral content, using dorsal-palmar radiographs. A study by Williams et al. (1991)
evaluating the capabilities of noninvasive techniques to estimate bone mineral content
(BMC) and bone strength of the third metacarpal in cattle demonstrated a correlation
coeflicient ( r) between BMC and radiographic photometry of .967 (P < .0001). This
valuewasshownto behigherthanrvaluesbetweenBMCandtwo othernoninvasive
methods evaluated, photon absorptiometry (.908, P < .0001) and ultrasonography (.406, P
< .0001). Studies have shown that RBAE ofthe third metacarpal is an adequate means of
demonstrating the changes in bone mineral content that occur during normal bone growth
(El Shorafa et al., 1979; Meakim et al., 1981; Frey et al., 1992) and exercise (Nielsen et
al, 1997). However, RBAEs determined by the above method can only indicate cturnges
in the section ofbone with the greatest optical density. A method that can measure total
changes occurring in the third metacarpal may be more valuable.

Single photon absorptiometry has been shown to be efi'ective in determining bone
mineral content of the third metacarpal (Jefl'cott et al., 1987). However, disadvantages of
this method include the necessity ofsedating horses in order to get the most accurate bone
scans, which take 60 3 each to complete, and preparing the horse and performing the
procedure requires approximately 30 min per horse. Therefore, this method is impractical
for utilization in a research setting when large numbers of horses are involved.

A method for determining cross-sectional area of the third metacarpal using
ultrasonography has been reported (McCartney and Jefi‘cott, 1987). In contrast to single

photon absorptiometry, ultrasonography has been reported to be a relatively simple and

23

reliable technique for measuring changes in bone. However, ultrasonography has a
limitation in that the changes that it measures are volumetric in nature.

These studies suggest that the use of radiographic photometry to estimate bone
mineralcontent ofthethirdmetaearpalinhorsesmaybethemost eficientandefl‘ective

method in a research setting.

24

CHAPTER 2

MATERIALS AND METHODS

Management of Animals

Sixteen Arabian yearlings, ll geldings and five fillies, with a mean age of 18.6 mo
(range 16.6 to 19.9), were pair-matched by sex and age and randomly placed into two
experimental groups. One groupwashousedin3.0mby3.4mboxstallswhilethe
second group was mintained on a 28,327 m2 pasture. The project was conducted during
thewintermonthstominimizedietarydifl‘erencesthatmayhaveoccurred ifgrazinghad
been available to the pastured horses. Each horse was fed 1.8 kg of a commercial
concentrate (Strategy for Performance or Breeding Horses, donated by Purina Mills,
Lansing, MI) per d, divided into two equal feedings (0600 and 1800), and had ad libiturn
access to mixed alfalfa-grass hay, satisfying NRC (1989) recommendations for long
yearlings and two-year-olds (Table l). The horses were maintamd on this diet through
the duration of the study. Based on 24-hr hay intake trials conducted every 28 (I, each
stalled horse consumed an average of 10.9 kg and each pastured horse an average of 9.8
kg of roughage per d. The horses were vaccinated against rhinopneumonitis, tetanus,
influenza, and equine encephalomyelitis. In addition, horses had their hooves trimmed and
were dewormed routinely, and had their wolf teeth removed prior to the start of trarnrng

Table 1. Calculated analysis of
total ration on an as-fed basis

fitment ,_Uait_§_. Ration

 

DE Meal/kg 2.169
CP % 14.7
Ca % .75

P 0/9 . 29

 

«1..— - -- -.l‘l." ' ‘ F'lw-a.‘ .747:

25

TrainingofAnirmls

TheprojectwasdividedmmmOphasesconsisfingofanM-dpre-fiainmgpefiod
anda56-dtrainingperiod. Duringthepre-trainingperiod,pasturedhorseswereallowed
fi'eeaccesstoexercise,whilestalledhorseswerewalkedlhrperdonameehanical
walker. Six horses fiom each group were randomly selected to enterthe training phase of
theproject. Onehorsefi-omthepasturedgroupwasremovedfi'omtheprojectafierd84
due to an accidental injury, hence only five horses were represented in the pastured group
dmingtheSG—daytrainingperiod. EachhorsewasriddenSdperwk,followingthe
nnrningfeeding,andthestalledhorseswerewalkedlhronamechanicalwalkeronall
non-ridingdays.

Dmingthefirstwkofthe56-dtrainingpefiod,thehorseswereinfi'oducedtoa
saddleandrider,gentledtoride,andtaughttoguideinaroundpenofl4.0mindiameter.
Duringthisstage,thehorseswereaskedtowalk,trot,andcanterinbothdirectionsforZO
to30minperd,thelengthoftimedependingontheprogressofthehorse. Riderswere
randomly assigned to horses (withinarider's capability). During the second through the
forn'thwkofthetrainingperiod,thehorseswereriddenineithera30.5mby6l.0m
outdoorarenaor 22.9 mby 54.9mindoorarena. Some of the horses were occassionally
ridden for a brief period of time in the round pen and subsequently ridden in the outdoor
orindoorarena. Asduringthefirstwkoftraining,thehorseswereaskedtowalk,trot,
andcanterinbothdirectionsfor20t030minperd. Horseswereriddenona768m
exercisetrackSdperwkfortheremaining4wkofthe56—dtrainingperiod. Eachhorse
was trotted one lap to warm up, galloped three laps, and trotted a second lap in order to

cool down. Average training distances per horse per d during this 4-wk stage were 1378

26

matthetrotand2146matthegallop.

Ond 122and 131,thehorseswereriddenintheindoorarenadueto inclement
weatherandpoor track conditions. Horseswereexercisedsimilardistancesinthe indoor
arenaond122and131asontheexercisetrackduringthelast4wkofthe56—dtraining
period. Ond 87, 94, 108,and 115,thehorseswereriddenintheroundpento

accomodate a behavior research project involving the same horses (Rivera et al., 1997).

Sample Collection

Dorsal-palmar radiographs of each horse’s left fi'ont leg were taken to determine
RBAE ofthe lateralandmdialcortices ofthethirdmetaearpal. Blood sampleswere
takenviajugularpunctmeat ll hrpost-feeding (Lepageetal, 1991) to determineserum
osteocalcin, 25-hydroxyvitamin D (V it D), parathyroid hormone (PTH), and serum Ca and
P concentrations. Twenty-four-hr urine collectiom were taken fi'om four randomly
chosen geldings fi'om each treatment group to measure urinary deoxypyridinoline,
pyridinoline, and minary Ca and P concentrations. Body weight, height at withers, third
metacarpal circumference, and body condition score (I-Ienneke et al., 1984) were recorded
for each horse every 28 d. Horses were weighed on a livestock scale. Additionally, 24-hr
intake trials were conducted every 28 d to determine the average kg of hay consumed per
(1 per stalled or pastured horse. On d 14, 42, 70, 98, and 126, horses from each
experimental group were randomly selected to lmdergo visual observation to determine
the number of elective strides taken, and in what gait, during a 24-hr period. Blood
samples were collected every 14 d, while radiographs were taken and urine collections

were conducted every 28 d.

27

Urine Collections

Twenty-four-hr urine collections were conducted on d 0, 28, 56, 84, 112, and 140.
Eight geldings randomly chosen for collection were allowed to become accustomed to
wearingasaddlepad, collectionharness, andsurcinglepriortothefirst24—hrcollection
ond0. Horsesweretiedinindividualstalls, withinreachofahayfeederandfeedtub,
throughthe dmtionofeachcollection. Urinewascolleetedinarubberbustire innertube
hung beneaththe sheathofeachhorse byacollectionharnessand surcingle. Urinewas
emptiedfromtheinnertubeatregularintervalsto decreasethelikelihoodoflosinga
sample or contamination. The horses were monitored during the entire 24 hr and were
allowed to drink water periodically throughout each collection. After 12 hr of each 24-hr
period,a 10% sampleofurinewasmeasuredandretainedforeachhorse. After
completion of each collection, another 10% sample of urine was measured and retained
fi'omtheammmtofurinecollectedduringthelatter 12 hr. Thetwo 10%urinesamples
for eachhorse were mixed thoroughlyand fi'omthat afinalurine sample for eachhorse
was taken at each collection, stored in a 250 ml bottle, and frozen for analysis following
the completion of the project. The total volume of urine excreted during each 24-hr

period was recorded for each horse.

Elective Exercise Observations
Observations of daily elective strides were conducted on d 14, 42, 70, 98, and 126.
Two horses fiom each experimental group were randomly selected to be visually observed
for three l-hr periods on observation days. The three periods were scattered throughout

theday,thefirstperiodinthemoming(between0600and0900),secondperiodinthe

28

afternoon(between1500and1800),andthirdperiodintheevening(between2100and
2400). Thenumber of elective strides takenperhorse,andinwhatgait, werevisually
observedandrecordedbyhand. Stridenmnbersfi'omthethreetimeperiodswerepooled
by gait for eachtreatrnent group, divided bythe number of horses observed (2), and
divided againbythenumber of observation periods (3). Thisnumberwasthenmultiplied
by24hr,or23hrinthestalledgroup,todeterminetheaveragenumberofelectivestrides
takenperhorseatthewaflefiotandcanterontlmtobservationday. Thestalledhorses
were also visually observed for 1 hr onthe mechanical walker on each observation day.
Stridenumbersfi'omthattimeperiodwerepooledbygaitanddividedbythenurnberof
horses omerved (2) to calculate the average number of strides taken per stalled horse at
thewalk,trot,andcanterfor1hronthemeehanicalwalkeronthatobservationday. The
datacoflectedforaflobservafionsdayswerethenpooledbygaitandn‘eaunemm
determine the average number of daily elective strides per horse, based on a 24-hr period
forpasturedhorsesanda23-hrperiodplusal-hrperiodonthemechanicalwalkerfor

stalled horses.

Sample Analysis
Dilutions were rmde of serum and urine for determination of serum osteocalcin,
urinary deoxypyridinoline and pyridinoline, serum and urinary Ca, and serum P. Serum
samples fordeterminationofVitDandPTHandurine samples formeasurementof
urinary P did not require dilution. Serum osteocalcin and urinary deoxypyridinoline and
pyridinoline concentrations were determined by competitive enzyme immunoassay

procedures (Metra Biosystems, Inc., Mountain View, CA). The results obtained fi‘om

29

deoxypyridinoline and pyridinoline assays were corrected for variations in urine
concentration by dividing the deoxypyridinoline or pyridinoline value (nM) by the
creatininevalue (mM) ofeachsample. FinalvalueswereexpressedasnM
deoxypyridinoline or pyridinoline per mM creatinine. Concentrations of creatinine in the
urine were determined by colorimetric assay (Sigma-Aldrich, St. Louis, MO). Urinary
deoxypyridinolineconcentrationswerealso determinedonaperdbasisbymultiplyingthe
concentrations obtained from the deoxypyridinoline assay by the total volume of urine
excreted per horse during the 24-hr period of each urine collection, and then multiplying
by the molecular weight. Radioimrnunoassays were used to determine Vit D and PTH
concentrations in the serum (INCSTAR Corporation, Stillwater, MN). Due to limited
funding, enough assay kits were purchased only to determine whether differences
appearedtoexistbetweentreatrnentgroups. Hence, notallsampleswereanalyzed
Coeficient of variation (CV) for assays used were determined using lmknown samples
provided by the company. Serum and urinary Ca were analyzed by atomic absorption
using a Smith-Hieftje 4000 (Thermo Jarrell Ash, Franklin, MA). Concentrations of P in
the serum and urine were determined using a DU 7400 spectrophotometer (Beckman,

Fullerton, CA).

Radiographs
Dorsal-palmar radiographs of the left third metacarpal were taken to determine
RBAE values using the technique descn'bed by Meakirn et al. (1981). An aluminum
stepwedge penetrometer was exposed simultaneously with each radiograph for use as a

standard of reference, necessary for the comparison of radiographs. Dorsal-palrmr views

30

weretakenwiththecasseflepositionedparallelmthecranialsmfaceofthe legand
centeredmidwaybetweenthepmrdnmlanddistalendsoffllethh'dmetacarpal Medial-
lateralviewsofthe third metacarpalwerenot includedinthesample collectiondue to
insuficient financial support.

A Bio-Rad Model GS-700 imaging densitometer (Hercules, CA) was used to scan
theradiographs transverselyat the nutrient foramenofthethirdmetacarpal. Alogarithmic
regression, developed using the thickness of the steps on the stepwedge, provided a means
to estimate bonemineralcontentinRBAE atthemardmumopticaldensityreadingofthe
lateral and mdial cortices.

The scansofthe dorsal-palmarradiographofthethirdmetacarpalandthe
aluminum stepwedge penetrometer were used to measure total RBAE in mm2 Al. The
area under the stepwedge curve corresponding to the steps with thicknesses of 14, 17, 20,
23, and26mmAlandthetotalareaunderthecurveofthethirdmetacarpalwere
determined. The total area under the five steps from the stepwedge scan was 1270 m2.
The total RBAE wasthendetermined bymultiplying thetotalareaunderthe curve ofthe
thirdmetaearpalby1270mm2Alandthendividingbythearealmderthestepwedge

curve.

Statistical Analysis
Physical characteristics, biochemical bone markers, and bone mineral content data
were analyzed as repeated measures using PROC MIXED analysis (SAS, 1997).
Treatment means, pairwise differences between the means, standard errors ofthe means

and differences, and t-tests indicating statistical significance of the means differences were

31

obtainedbyinclusion of the LSMEANSstatementintheanalysis(SAS, 1997).
Treatment, day, andday‘treatment were includedinthemodel to determinetheefl‘ects of
stallingonnormalbone growth. Difl‘erenceswereconsidered significantatP<.05. When
difieremeserdnedbetweenthefiafledandpastmedgroupsmthemdmepmjeafio
values(baseline)foreachhorseweresubtractedfi'omallothervaluesforthatanimaland
analysiswasperformed onthechange fiom baseline. Results were graphedwiththeSEM

indicated for each mean.

32

CHAPTER 3

RESULTS

Weight, Height, Third Metaearpal Circumference, and Body Condition Score
No treatment differences occurred in body weight, height at withers, third

metacarpal circumference, or body condition score. However, horses kept in stalls had
higherbodyweightatd140thanhorsesmaintainedonpasture(P<.05). Stalledhorses
began the project at an average weight of349 : 10.2 kg and increased to 405 :11.1 kg
byd140 foranaverage gain of56 kg. Pastured horseshadaninitial weight of342: 10.1
kgandincreasedto 367: 11.3 kgbyd 140 foranincrcaseof25 kg. Bodyweight for
bothtreatmentgroupsincreasedsteadilyfiometo d84andremainedrelativelysteady
throughtheduration ofthetrainingperiod to d 140. Mean height forthe stalledhorsesat
d0was 143 j; 1.0 cm, increasing 3 cmto 14611.5 cmbyd 140. Pasturedhorsesalso
beganthe project at an average height ofl43 $1.0 cmand increased 3 cmto 146 i 1.6
cmbyd 140. Third metacarpal circumferencewas 18.4:04 cmatthe startofthe project
and 18.6:03 cmbyd 140 inthehorseskeptinstalls. Pasturedhorseshadaninitialthird
metacarpal circumference of 18.6 i 0.4 cm compared to 19.0 j; 0.4 cm by d 140. Body
condition score (BCS) rennined relatively unchanged for both treatment groups. Stalled
horses began the project at an initial BCS of 6.2 1: 0.1 and increased 0.5 to a BCS of 6.7 i
0.2 atd 140. Bodycondition score forthepasturedhorseswas6.1 10.1 atdOand6.4j;
0.2 at d 140 for an increase of 0.3. Body condition score for both treatment groups
decreased fi'om d 84 to d 112, as horses entered training, and subsequently increased

through d 140.

33

Elective Exercise
The number ofstrides per day ofeach gait, calculated fi'om visual observation,
indicated large variations between treatment groups in elective physical activity (T able 2).

Table 2. Calculated daily elective strides per horse fiom observation.

 

 

 

., (Walkmmthrot Canter , Total
Pastured
24 hr on pasture 4076 389 674 5139
Stalled
23 hr in stalls 804 0 0 804
1 hr on mechanical walker 3171 55 10 3236
Total for 24 hr 3975 55 10 4040

 

Radiographic BoneAluminumEquivalence

Analysisoftheradiographsshowedafiendtowardadifierencebetweenthetwo
treatmentgroupsatdOinRBAEofboththelateral(P<.15)(Table3)andmedial
cortices(P<.1) (Table4) ofthethirdmetaearpal Therefore, baselinevaluesforlateral
and medial RBAE for each horse were subtracted from RBAE values for the appropriate
cortexonallotherdaysforthatanirnal. AnalysisofthechangeinlateralRBAEfiome
(Figure 1) showed that stalled horseshad lower lateral RBAEthanpasturedhorsesatd28
(P<.05), whichremainedlower through most of the duration of the project,whilehorses
nuintained onpasturehadincreasing lateral RBAE(P<.05). Pasture-reared horseshad
greaterlateralRBAE atd28,56,and140(P<.05),andhadatendencytobegreateratd
112(P=.07). ThechangeinRBAEofthemedialcortex(Figure2)tendedtobedifi‘erent
betweentreatments(P=.1). HorseskeptinstallshadlowermdialRBAEatd28(P<
.05), which tended toremainlowerthanhorsesmaintainedonpasturethrough most of the

project(P= .1). MedialRBAEs were greateratd28 (P<.05)andtendedtobegreaterat

34

Table 3. Means table for lateral RBAE (mm Al).

 

 

DaL_0 _ 28. _____. A WM- 8.4W_-__--_.,__1..1,2.______,.. 140
Stall 20.4231 19.495 19.499 20.062 19.843 19.841
SEM 0.313 0.315 0.304 0.425 0.415 0.269
Pasture 19.770 19.646 19.908 20.208 20.339 20.189
SEM 0.313 0.315 0.315 0.442 0.446 0.293

 

fintableindicatespasturedifi‘erentthanstalledatgivenday(P<.15)

Table 4. Means table for mdial RBAE (mm Al).

 

 

_pay_____9§ W _, 28c 56be 84‘” 1121|" 140'
Stall 21,363+ 20.335 20.495 21.608 20.957 21.501
SEM 0.215 0.280 0.295 0.601 0.416 0.185
Pasture 20.788 20.566 20.938 21.214 21.762 21.533
SEM 0.215 0.280 0.308 0.629 0.450 0.202

 

fintableindicatespasturedifi‘erentthanstalledatgivenday(P<.1)
“Days lackingacommonsuperscriptdifl’er(P<.05)

2 1.5- —u— Pasture“

 

 

 

 

a l o- —o— Stall
g 0.5-1 ,, '1' . at
a:

- 0.0 ‘ 3 ‘7' I l

i 1 - x

.3 -0.5-

‘s III- . .

g. .100‘ . . l

H

.2

Q '105‘ I l r r l
0 28 56 84 112 140

Figure 1. ChangeinLateral RBAE (mmAl) versus day ofproject.
*inlegendindicatespasturedifi‘erentthanstalled(P<.05)
*ingraphindicatespasturedifi‘erentthanstalledatgivenday(P<.05)
fingraphindicatespasturedifi‘erentthanstalledatgivenday(P=.07)

35

 

 

 

 

2 15‘ —u— Pasture? r

s. l 0- —o— Stall _ 1'

g 0.5q W "

a . .  ,

% 0.0- 3 v ‘ r

E -0.s- :: .

.5 1

o O

a” -100- o

'2 l 5

U ' ° ' I I I T I
0 28 ° 56 3° 84 a 1123‘" 140 3"

Day

Figure 2. Change inMedial RBAE (mm Al) versus day of project.
finlegendindicatespasturedifl‘erentthanstalled(P=.1)
*ingraphindieatespasturedifi‘erentthanstalledatgivenday(P<.05)
fingraphindicatespasturedifi‘erentthanstalledatgivenday(P<.l)
“Days lackingacommon superscript difl‘er(P < .05)

d56andd112(P<.1)inhorsesmaintainedonpasture,comparedwithhorseshousedin
stalls. Radiographic bone aluminum equivalence of the medialeortex increased fi'omd 28
to d 112. Notreatmentdifi‘ereneesoccurred intotalRBAE ofthethirdmetaearpal(Table
5). For both treatment groups, total RBAE remained relatively unchanged throughout the

project, except for a slight decrease fi'om d 112 to d 140.

Serum Osteocalcin
No treatment differences occurred in serum osteocalcin concentrations (Figure 3),
however serum osteocalcin was lower in the stalled horses at d 14 when compared to
horses maintained on pasture (P < .05). Following d 14, osteocalcin concentrations in the

stalled horses returned to baseline. A high degree of variation in osteocalcin

36

Table 5. Means table for total RBAE (mm2 Al)
- pay 0‘” W28"? , - 1 56': 84'” 112' 140”

 

 

 

 

Stall 526.808 486.477 523.633 546.850 539.985 500.840
SEM 27.159 35.500 30.032 83.375 52.816 35.497
Pasture 491.257 505.700 574.647 528.663 577.435 482.279
SEM 27.159 35.500 30.912 85.882 56.353 38.496

”Days lacking a common superscript difl‘er (P < .05)

concentrationspresent inthe serumofpasturedand stalledhorsesoccurred duringthe
latter half of the project. Stalled horses tended to have greater osteocalcin tlnn pastured
horses at d 98 (P < .1), while pastured horses indicated a trend toward greater osteocalcin
concentrationsthanhorses housed installsat d 140 (P < .1). Serumosteocalcindecreased

inbothtreatment groups fiomd 112 to 126, followedbyanincrease throughd140.

Urinary Deoxypyridinoline and Pyridinoline

Urinary deoxypyridinoline concentrations based on creatinine (Figure 4) were
greateratd28 inhorseshousedinstallsthaninhorsesmaintamdonpasture(P<.01).
Similar results occurred in urinary deoxypyridinoline concentrations based on total daily
urinary output (P < .1) (Figure 5). Following d 28, deoxypyridinoline inthe stalled horses
returned to baseline. No other treatment or time difl‘erences occurred in urinary
deoxypyridinoline. Urinary pyridinoline concentrations (Table 6) appeared to decrease in
both treatment groups through d 84, and subsequently increase slightly through d 140. No

treatment differences occurred in pyridinoline concentrations.

37

i
.,
a
E

45q —o-— Stall

Osteocalc ,ng/ml
8 3 3
I I I
.4.
V.
.I
l—;,’>

 

 

25- ‘r
20- '

I
15.

 

63" ifb‘z'sabiz abs?”“rl1%“"9ir“°df"ris°130“

Day

Figure 3. Serum osteocalcin (ng/ml) versus day of project.
I"ingraphindicatespasturedifi‘erentthanstalledatgivenday(P<.05)
tingraphindicatespasturedifl‘erentthanstalledatgivenday(P<.l)
““Days lacking acommon superscript difl‘er (P < .05)

H
i—r
I

in
\D G
I I

 

a \l
I I

 

Urinary Dpd, nM/mM
‘f

 

 

 

84 112 150

a.
8:

Day
Figure 4. Urinary deoxypyridinoline (nM/mM) versus day of project.

" ingraphindicatespasturedifl‘erentthanstalledatgivendaye<.01)
n=4foreachtreatmentgroup

38

0.45 -
0.40 d

 

0.35 -

N’s/d

9 0.30-

0.25 -

Urinary

0.20 ‘ q h

 

 

0.15.1 , T

 

 

c.
E
g
E

Figure 5. Urinary deoxypyridinoline (g/d) versus day of project.

I
112

1
140

fingraphirrdicatespasturedifl‘erentthanstalledatgivenday(P<.l)

n=4foreachtreatmentgroup

Table 6. Mean table for urinary pyridinoline (nM/mM).

 

 

Day 0' 28‘ 56‘ 84" 112'” 140‘
Stall 263.266 219.822 180.592 145.179 144.456 184.594
SEM 39.386 36.616 23.109 15.507 21.898 24.571
Pasture 155.831 156.248 153.397 123.926 158.813 174.615
SEM 39.386 36.616 23.109 14.202 21.898 24.571

 

“Dayslackingacomnnnsuperscriptdifl‘er(P<.05)
n=4foreachtreatmentgroup

39

Serum Minerals, Vitamins, and Hormones

No overall treatment difi‘erences occurred in serum Ca concentrations (Table 7).
Inboththe stalledandpasturedhorses, serumCaappeared to increase slightly fiometo
d 14, followed by a decrease through d 84, remained relatively stable up to d112 and then
increased through d 140. A high degree of variation occurred between treatments in
serum Ca concentrations during the latter halfof the project. At d 84, horses housed in
stalls tended to have greater serum Ca (P < .1), however pastured horses had greater
serum Ca at d 98 (P < .01). Following d 98, concentrations ofserum Ca did not differ
between treatment groups. Serum P concentrations (Table 8) were lower in the stalled
horses atd0whencompared to horsesmaintainedonpasture(P< .01), andrermmd
lower through most of the project (P < .05). Pasture-reared horses had greater serum P
concentrations at d 0, 28, 112, and 140 (P < .01), and at d 70 (P < .05), however stalled
horseshadgreater serumP at d 98 (P < .05). SerumP mcreased inbothtreatment groups
fiom d 0 to d 14 and then decreased below baseline values by d 28, increasing slightly
through d 84. From (1 84 to d 98, serumP decreased markedly, changed little through d
126, and then increased through d 140. No treatment difl‘erences occurred in serum Vit D

or PTH concentrations (data not shown).

Urinary Minerals
Noueatmemdifl‘erencesbetweenstafledandpastmedhorsesoccunedmurinary
Ca (Table 9) or P (Table 10) concentrations. In both treatment groups, urinary Ca
decreased fi'om d 0 to d 84 and then increased through d 140 to concentrations Similar to

those found in the first half of the project. Urinary P concentrations in stalled and

40

Table 7. Means table for serum calcium (mg/d1).

 

 

 

 

 

 

 

 

 

 

 

 

 

Day 0' 14° 28° 42' 56" 70°‘
Stall 12.778 13.275 13.067 13.244 11.926 11.317
SEM 0.273 0.241 0.440 0.268 0.425 0.254
Pasture 13.188 13.484 13.350 12.793 12.111 11.265
SEM -- 0.273 0.241 __ 0.440 0.268 0.425 0.254

Day 84° 98°°° 1 12“° 126bc 140bc
Stall 11.127? 10.446" 11.051 11.575 11.902
SEM 0.306 0.295 0.267 0.247 0.403
Pasture 10.396 11.588 10.601 11.230 11.230
SEM 0.306 0.295 0.292 0.271 0.441

”in table indicates pasture different than stalled at given day (P < .01)
‘l’ in table indicates pasture difl‘erent than stalled at given day (P < .1)

M°Days lacking a common superscript differ (P < .05)

Table 8. Means table for serum phosphorus (mg[dl). . _ .

98);. 0°. 1. If. I 28:“. ”12%“... .. - 5A,.-. ZQ'I.___
Stall“ 4.829“ 6.216 4.471” 5.499 5.371 5.371“
SEM 0.192 0.251 0.142 0.184 0.202 0.265
Pasture 5.635 5.964 4.996 5.370 5.264 6.138

SEM 0.192 0.251 0.142 0.184 0.202 0.265
Dax__ __ ,_ 34'” 934 , . 1.12:. 12:6“ I140“-__.
Stall 5.635 5.072" 4.096" 4.605 4.951”

SEM 0.267 0.293 0.135 0.238 0.344
Pasture 6.020 4.123 4.956 4.678 6.959

SEM 0.267 0.293 0.146 0.261 0.377

 

*inheadingindicatespasturedifi‘erentthanstalled(P<.05)

** in table indicates pastlne diflemnt than stalled at given day (P < .01)
* in table indicates pasture different than stalled at given day (P < .05)
‘MDays lacking a common superscript differ (P < .05)

41

pastured horses appeared to follow each other in a slight increase throughout the project,
althoughurinaryPinthepasturedhorses increasedto agreaterdegreethaninthestalled

horses following (1 112, creating a tendency for pastured horses to have greater urinary P

atdl40(P<.l).

Table 9. Means table for urinary calcium (g/d). , ,

 

 

 

 

 

Dar- - - 0° ‘ 281A. A" - - 84° 1 12°? - 1402,. __
Stall 0.446 0.582 0.303 0.154 0.308 0.372
SEM 0.122 0.210 0.062 0.032 0.086 0.120
Pasture 0.677 0.704 0.334 0.262 0.456 0.589
SEM 0.122 0.210 0.062 0.032 0.086 0.120
“Days lacking a common superscript difi‘er (P < .05)
n = 4 for each treatment group
Table 10. Means table for urinary phosphorus (g/d).
Day--...-_-_-0d, 2871 56f,“- ,- 84'”- 112“ 140'
Stall 62.222 143.638 122.266 159.413 124.478 11 1.8411

SEM 21.053 61.161 46.499 76.238 44.237 113.857
Pasture 56.928 41.520 129.003 313.308 119.228 420.558
SEM 21.053 61.161 46.499 76.238 44.237 113.857

 

‘l' intable indieatespasturedifi‘erentthanstalledatgivenday(P< .1)
““Days lacking a common superscript differ (P < .05)
n = 4 for each treatment group

42

CHAPTER 4

DISCUSSION

A horse’s skeletal system will reach firll maturity at approximately 4 yr of age (El
Shorafa et al., 1979). Lawrence et a1. (1994) reported that maximum bone mineral
comemandbreakingsuengtharemtreachedintheequinethirdmetacarpalumfl
approximately6yrofage, andmaximumbreakingloadisnotreacheduntilaroundS yrof
age. The most significant changes in the cross-sectional area and inertial properties of the
third metacarpal bone, indicative of the bone’s resistance to compressive stress and of the
bendingstrengthofthe bonebyalterationsinmassarormditscentralaxis, respectively,
have been shown to occur between a Thoroughbred’s yearling and 2-yr-old yr
(Nunamaker et al., 1989). El Shorafa et al. (1979) demonstrated that in the horse, bone
mineralcontent increases sharplythroughthe firstyearoflife, and subsequently increases
filrtherupuntilapproximately7yrofage, butatamuchslowerrate. Thesestudies
indicatethat the skeletal systemofthe young, growing horse isverydynamic and hasthe
capability of developing to its full genetic and functional potential During growth, the
modeling and remodeling mechanisms are important in the development and maintenance
ofan optimal skeletal structure for a young horse in training and competition.

Ifthe fimctionalstrainonthelimbsissuddcnlyreduced, asinbedrest(LeBlancet
al., 1990) or immobilization (Kannus et al., 1996), the skeletal architecture responds by
lowering bone mass to a level appropriate for the new strain (Lanyon, 1984). Results of
thisstudysuggestthathoushgyearlhghorsesinstaflswithhmitedaccessto exercisemay

negatively affect bone growth compared to that experienced by yearlings allowed to

43

remainonpasture,thuspredisposingtheformergrouptoinjurywhenhigherstrainrates
areintroducedduringtrainingorcompetition. Thesenegativeefi‘ectsasindieatedby
analysis of RBAES, osteocalcin concentrations, and deoxypyridinoline concentrations,
were most apparent between l4and28dofplacement ofyearlingsinstalls.

Nielsen et al. (1997) determined that fewer skeletal-related injuries occurred to
yomghomesmmnialnammgwhenRBAEswemgreatcrmmenndialandlateralwrfices
ofthethhdmetacarpaLcomparedwfihthepahnarcofleLattheeommemememd
training. ThemajorityofchangestothebonemineralcomeMOfthethirdmetacarpalin
responsetotrainingoccurredinthemedialandhteralcortices. Basedonthefindingsof
thisstudy,esthmtedbonemmeralmmemmthepmsem$tudywasdetemmedforonly
themedialandlateralcorticesofthethirdmetacarpal.

Radiographic bone aluminum equivalencies have been highly correlated with bone
mass(Meakimeta1.,1981),which,inturn,hasbeendescribedasthebestmeasmable
determinant ofbonestrength(Kimmel, 1993). Thelower RBAEs ofboththemedialand
hteralmnicesofmethhdmaacmpalinthestafledhorsegwmpmedwhhthepastmed
horses,atd28oftheprojectindicatednotonlylossofbone,butalsolossofbone
strength. ThisissupportedbytheresultsofastudyconductedbyNielsenetal.(1997)
examiningthechangesinskeletalstrengthof53 QuarterHorsesputintoracetrainingat
18 mo of age. Using atechnique correlating RBAE and total bone mineral content
(NielsenandPotter, 1997), Nielsenetal. (1997) demonstratedadecrease inRBAE of the
totalcross-sectionalareaofthethirdmetacarpalsoonafiertheonsetofracetraining. The
lowestRBAEswereattained60to 100daysafterthestartoftraining,atwhiehtimespeed

was introduced into the training program and the highest number of skeletal-related

injuriesoccurred. Priortothestartoftraining,thehorseswerekeptonpastmewithfiee
access to exercise. Upon commencement oftraining, all horses were maintained in stalls
atalltimesexceptwhentheywerebeingexercised. Itwasnotdeterminedwhetherthe
decreasehbonemheralewastheresuhofbonemmodefingcausedbymcreased
sfiainratesonflnboneassocifledudthuammgormemsuhofbomnndelmgresulfing
fiomdecmasedsfiahratesassochtedwfihmechmgemhousmgfiompastumtostaflsat
thestartoftraining.
AhhoughRBAEmeasmementsinthepresemSMydidnotindicatediflerencesin
totalbommmeralwmembeMeenmestafledandpastmedhorsesifmeremhsobtamed
byNielsenetal.(1997)wereinresponsetothechangeinhousingmoresothantothe
mmnemememOfUainnghesestudiessupponthehypothesisthatfimifingayomg
horse's access to flee exercise will result in negative efl‘ects on bone development. Horses
werefiansfenedfiompastmetostallsandplacedhtofiainingatdOofthepmject
conducted byNielsen et a1. (1997). The lowest total RBAEs ofthe third metacarpal did
notappeartooccuruntil60tolOOdaflerthecommencementoftrainingandchangein
housing. However, after d 0, a second set ofradiographs was not taken until d 64 ofthe
project. Therefore, we do not know if the lowest RBAEswereactuallyattainedprior to d
60,asinthepresentstudy,inwhichthelateralandmedialcorticesofthethirdmetacarpal
anamedthelowestRBAEswnhm28dofstaflmgandsubseqmmlyretmmdtobasem
afierd28. Inthepresentstudy,latcralRBAEsremainedlowerinthestafledhorsesthan
inthehorsesmaintainedonpasturethroughthedurationoftheproject. Similarly,medial
RBAEs tended to remain lower in stall-reared horses when compared with pastured horses

thoughmostoftheproject. BothlateralandmedialRBAEswereunafi‘ectedbythe

45

commencementoftrainingatd84. Alowerintensityconditioningprogramwasusedin
thefirsttw028-dperiodsoftraininginthestudybyNielsenetal.(1997),whencompared
withthefiahingregimenusedinflreS6—dfiainingperiodinthepresem8tudy. Therefore,
thedecreascmbommmeralwmemOfthethhdmemcarpaldemonsumwsoonafierthe
onsetofracetraininginthestudybyNielsenetal.(l997)wasmostlike1ytheresu1tof
bone modeling due to decreasedstrainsonthebone associatedwiththechangeinhousing
fi'ompasturetostallsatthestartoftraining. Theapparentlongerperiodofboneloss,
comparedtothatinthepresentstudy,mayhavebeentheresultofthehorsesreceiving
somemechanicdsfiahonthebad—bearhgbomsassociatedwfihhainmg,causinga
slowdownintherateofboneloss. Themmatedecreaseinbonemineralcontentinthe
stafledhomesmthepresemstudyismtwpfismgwnsmmathousmgwndhiom
weretheonlytreatmentdifl‘erencesatdOoftheproject.
AstudywasconductedattthniversityofFloridalookingattheefl‘ectsofracc
WonbonemmeralwmemOfthethhdmetacarpalmyoungThomughbredhom
lmder field conditions (Porr et al.,1997). Two treatment groups were designated by
difl'erencesinmaturity. GroupAconsistedofhorsesvisuallydeterminedtobennre
maturethanhorsesinGroupB. Ascoltsarenormallylargerinbodysizethanfilliesat
around2yrofage,GroupAwasmadeupof11coltsandonlySfillies. GroupB
consistedofhorsesthatappearedtobelessdevelopedthanhorsesinGroupA,and
thereforecontained8filliesandonly5colts. Themeanageofthehorsesatthe
commencementoftrainingwasyoungerinGroupAthaninGroupB. Inadditiontothe
uneven numbers of male and female horses in each group, primaryhousing conditions

difi'eredbetweengroups. Fillieswereplaced,asagroup,intoalargcpastme,whilecolts

46

were paired by temperament and placed into snmll paddocks. Investigators in that study
comludedflmtthedeereasembonemineralcomemobsewedemupAshorflyafier
commemementoffiainmgwastheresuhofhorsesmeupAbeginnmguainingatan
earlieragethanGroupB,whichdemonstratedanincreaseinbonemineralcontentupon
commencementoftraining. However,resu1tsofthepresentstudysuggestthatthe
difl‘emncesobsewedbetweengroupsmmemsponseofbommmeralwmemofthethhd
mumpalmtheonsaofuammgwasmeresuhofthedifl’eremesmtheprhmryhousmg
conditionsbetweengroups. GroupAconsisted ofmoremalehorseshousedinsrnall
paddocks,whileGroupBconsistedofmorefemalehorsesnnintainedinalargepastm'eas
agroup.
IncompafisontothestalbdhorsesmedialRBAEsmthepasnuedhorsestended
toincreaseinapatternofgrowth,andlateralRBAEsappearedtofollowasimilarpattern,
although daydifl‘erenceswere statistically insignificant forlateralRBAE. Duringnormal
bone growth, which, according to El Shorafa et al. (1979), is not completed until
approximately4yrof age inthe horse, bone formationisoccurring fasterthanbone
resorption(MaenpaaetaL,1988),resultinginagradualincreaseinbonemineralcontent.
ThehorseskeptinstallsshowedanincreaseinmedialRBAEfi'omd56throughd84(P<
.05),thusreturningto baseline. Thismayindicatethattheskeletalsystennofthestalled
horses 1nd adapted to a new, lower functional load and subsequently returned to a pattern
of bone growth appropriate for their optimal strain environment. After commencement of
fiammgmd84,nndthBAEsmstafledhorsesceasedmcreasmgandevenappearedto
decrease. Ifuaininglmdnotbeeninuoducedatd84,itisspeclflatedwhetherthestafled

horses would have eventually returned to a similar bone mineral content and growth

47

patternasthcpasturedhorses.
Noueannemdifi’ereneesoccurredintotalRBAEofthethh-dmetaearpaland,

exceptforaslightdecreasefi'omd112tod140forbothtreatmentgroupscombined,
remainedrelativelyunchangedthroughouttheproject. Theslightdecreaseatdll2may
hflicatemadapfiveresponsewuainmg,althoughIMeralandmedialRBAEsdidmt
follow this pattern ofdemineralization and therefore its relevance is skeptical.

Osteocalcin is a biochemical marker ofbone turnover synthesized predomimntly
by osteoblasts, or bone-forming cells (Price et al., 1980), and is thus considered a measure
of osteoblasticactivity(Price, 1982). Inthepresentstudy,lowerserumosteocalcin
wmenuatiommthestalledhomesatdusuppontheapparembonebssdemomuaed
byRBAEsasanindicationofaslowerrateofosteoblasticactivityresultingfi'oma
decreaseinphysicalactivity. Maenpaaetal.(l988)reportedasimilarimmediatedecrease
inosteocalcininfoalsfiansfenedfiompastmetostallsforthewhnermomhs Amarked
decreasemserumosteocabmwmennmiomoccmredwhhmlnnofstalfingmresponse
tothedhnhishedmechanicdsfiahonthehmbsofmefoalsassocimwwhhstaflmg.

Exceptforthelowserumosteocalcininthestalledhorsesatd14andthe
unexphmablevafiationmosteocalcmwmenfiatklmmboththepastumdandstalbd
horsesdmingtheS6—dfiainingpefiod,osteocakhcomenfiafiommflnserumofboth
treatmentgroupsremainedrelativelystable. Thisstabilitysuggestsasteadyrateof
osteoblastic cell activity considered normal in young horses experiencing growth and
development. Studies have shown that serum osteocalcin concentrations are high in
young horses and decrease with age, indicating a slowdown inthe rate ofbone formation

inmaturehorsescomparedto foals(Lepageetal., 1990). Regardlessofwhetherbone

48

formtion was due to bone growth by the modeling mechanism or the replacement of
irnnmture or damaged bone by the remodeling mechanism, osteocalcin concentrations
measured in the present study were relatively normal for long yearling/two-yr-old horses.
Lepage et aL (1990) determined that mean osteocalcin concentrations for female horses
less than 1 yr ofage and between 1.5 and 2.5 yr were 47.3 and 35.7 ng/mL, respectively.
The mean osteocalcin concentration for both treatment groups combined was 33.1 ng/mL
in the present study.

In contrast to osteocalcin, urinary deoxypyridinoline and pyridinoline are released
by the activity ofosteoclastic cells, or bone-degrading cells, and are thus considered
indicators of bone resorption (Gomez et aL, 1994). However, differences in the origin and
activity ofthese pyridinium cross-links may influence their comparative usefirhrcss as
markers ofbone resorption. Pyridinoline isprimarilypresent inboneandcartilagermtrhr
and deoxypyridinoline is present in bone collagen ofthe organic matrix (Delmas, 1993;
Robins et al., 1994). These crosslinks are released into circulation by the action of
osteoclastic cells as bone collagen is digested during the degradation process (Gomez et
al., 1996). Pyridinoline has also been reported to be released from sources otherthan
bone in response to other degenerative conditions (Robins et aL, 1994). Therefore, except
for a slight decrease in urinary pyridinoline concentrations in both treatment groups
throughd 840fthepresent study, followedbyaslightincreasethroughd 140, itisnot
surprising that no treatment difl‘erences occln'red in pyridinoline concentrations.
Additiomlly, the srmllnumberofhorses representedirreachtreatmentgroup (n=4)
during urine collections may have contributed to the lack of specificity of pyndmolme as a

markerofboneresorption. Adifferencebetweentreatmentsnnyhavebeenmore

49

apparent with larger sample numbers.

Urinary deoxypyridinoline concentratiom were greater in the stalled horses than in
thehorsesmaintainedonpastureatd28. Allowingthat deoxypyridinoline isareliable
mrkmofbomdegradmionresuhsmdimmmmalbdhorseswemexpefiemmgaperbd
ofincreasedboneresorption28 dafierbeingplaced into stalls. Theapparentdecreasein
bone formation activity at d 14, as indicated by low serum osteocalcin, leads to the
speculation of whether urinary deoxypyridinoline could have been at its highest
concentration when osteocalcin was at its lowest. Without urine samples at d 14 for
determination of deoxypyridinoline concentrations, this remains unanswerablc. Vico et al.
(1987) demonstrated an inverse relationship between bone resorption and bone formation
inhealthymale subjects voluntarilysubjectedto 120dofbedrest. Bone biopsydata
showedMadecmasemminerafizafionrateandasnnuhaneoushrcmasembonecefl
resorption activity occurred in completely irmnobilized subjects.

Ineomparisontothe stalled horses inthepresentstudy,pasturedhorses
maintained a relatively stable concentration of urinary deoxypyridinoline through the
duration of the project. Black et al. (1997) demonstrated a strong correlation between
both osteocalcin and deoxypyridinoline concentrations and skeletal growth patterns in
foals. Osteocalcin and deoxypyridinoline concentrations reportedly decreased as specific
growthrates, including bodyweight, girth, andwitherheight, graduallydecreasedfi'om
birth to 1 yr of age. Although deoxypyridinoline concentrations did not follow a similar
patternintheprwent study,thismayberelatedto difl‘erencesintheageofthehorses
participating in each study and a faster growth rate occurring in the yormger horses.

Stalled horses quickly equilibrated to the decrease in strain on the load-bearing

50

g.

4-:

bones, as demonstrated by low osteocalcin concentrations inthe stalled horses, as well as
highdeoxypyridinoline concentrations,retm'ningto baseline followingd 14andd28,
respectively. This equilibration suggests that an adaptive mechanism was triggered to
appropriately match the skeletal architecture to the new functional load (Lanyon,1987).
Thelowerloadsplacedonthebonedirectlyorindirectlyindicatedtothebonecells
mspomibkfortheadapfivemsponseMunmcessaryemrgywasbemgusedtonnhnam
andtransporttheinitialskeletalmssmubinandLanyon, 1985; Lanyon,1987). The
processofbonemodelinglowersbonemassrmtilanewoptimalstrainenvironmentis
reached(Lanyon, 1984).
Nooverafltreafinentdifl‘erencesoccmredinserumCaconcenfi'ationshrthe
pmsemstudy,althoughasfightmcreasemserumCaappearedtooccmmthestalbd
horsesfrometod14. TheimmediateincreaseinserumCaappearstoreflectthe
releaseofCafi'omboneintothecirculationdmingboneresorption. Thisissupportedby
high urinary deoxypyridinoline concentrations inthe stall-reared horses at d 28, indicative
ofahighrateofboneresorption,andthelowlateralandmedialRBAEsatd28,indicating
alossofbonemineralcontent. Krook(1968)hasdemonstratedanincreaseinserumCa
associatedwiththereleaseofCaintothech'culafiondmingboneresorption Nielsenetal.
(1998)alsoreportedaninverserelationshipbetweenserumCaandRBAEsdruinga
periodofboneresorptionintheearlystagesofatrainingprogram.
Foflowingd14,serumCaconcentrafionsinboththestafledandpastluedhorses
decreasedthhewmnemeMoffiahhgfidM,mmahedmhtivelystablethroughd
112,andthenincreasedthroughdl40. ThedecreaseinserumCathroughd84maybe

attributable to the deposition of Ca into bone from the circulating Ca pool in support of a

51

 

period of bone formation (Bronner, 1993). This makes sense in the horses maintained on
pasturebecausetheyappearedto beexperiencingapatternofskeletalgrowth,basedon
medialandlateralRBAEs. MedialRBAEincreasedinthestalledhorses fiomd56
through d 84 (P < .05), possibly indicating that they too were experiencing bone growth at
their new optimal strain environment, albeit at lower RBAEs than those found in the
pasturedhorses. BothlateralandmedialRBAEswereunafl‘ectedbythecommencement
oftraining, whichmayexplainthelackofactivityinserumCaconcentratiominboth
treatment groups through d 112 of the project.

SerumCaconcentrationsaremaintainedwithinanarrowrangethroughtight
regulation by the body (Krook and Lowe, 1964). A secondary function of bone
remodeling, in addition to an adaptive response to mechanical strain on the load-bearing
bonesandreplacement ofoldordamagedbone, istomaintainserumCalevels(Riggsand
Evans, 1990). Inadditionto bone, Cacanbetakenfromthedietandmuscletissuein
order to maintain conditions of mineral homeostasis in the blood (Bronner, 1993). Due to
the body's tight regulation of the Ca pool, large deficiencies in dietary Ca intake are
necessary to cause marked variations in serum Ca concentrations. In the present study,
serumCaconcentrationswerewithinthenormalrange forhorses(DuncanandPrasse,
1986) and diets were fed to satisfy NRC (1989) recommendations for the age and activity
ofthe horses. Therefore, the variation in Ca concentrations in the serum ofthe stalled and
pasturedhorseswerenot likelytheresultofCaintake.

Lepage et al. (1991) determined that serum phosphate concentrations follow an
inverse pattern to serum Ca concentrations, serum phosphate being low when serum Ca is

highbasedonamlysisofbothmineralsovera24hrperiod. Thispatternwasnot

52

observedinthepresentstudy. InfhctPconcenn'ationsinthesermnofstalledhorses
increaseddrannticallyfi'ometod14(P<.01),similartotheapparentincreaseinserum
Caconcentrationsduringthesamel4-dperiodinthestalledgroup. Thisincreasein
serluanayalsobetheresultofthereleaseofbonemineralintocirculationduringa
periodofboneresorption.

SenmrPcomenfiationshthestafledandpashnedhorsesweredifl'eMatdOJhe
stalled horses having lower P concentrations when compared with horses mintained on
pasture. SemmPconcenfiationsarenotastightlyregulatedasCaconcentrafionsinthe
sermthereforevariafionsindietaryhrtakewiflquicklyafi‘ectsemmP. Cymbalukand
Chfistison(1989)demonsfifiedthmfeedmgvaryingamomtsomethedietdnealy
influencesserumPconcentrationsinhorses. Forexample,feedingadiethigherinP
resultsinhigherPconcentrationsintheserum. Inthepresentstudy,thedifl‘erencein
serumPconcentrationsbetwecnthetwotreatmentgroupsatdOmaybereflectiveofa
reductioninhayintake,andthusadecrcaseindietaryPintake,bythestalledhorsesdue
tothestresscausedbythechangeinhousingfrompasturetostalls. Thehighdegreeof
variationinserumPconcentrationsbetweenthetwotreatmentgroupsduringmostofthe
projectmayalsobeattributabletovariationsindietaryPintakes. SerumPconcentrations
werewithinthe normal range for horses(DuncanandPrasse, 1986).

Urinary Ca output can accumulate flour a number of sources, including the diet,
muscle tissue,andbone (Bronner, 1993). Inhumansubjects, firstingurinaryCameasured
fiom a morning sample and corrected by creatinine excretion is useful in detecting a
notable increase of bone resorption activity (Delmas, 1993). Using urine samples from

24-hr urine collections, urinary Ca concentrations were easily determined in the present

53

 

study by atomic absorptiometry. Urinary Ca concentrations are highly variable, depending
on, for example, the age of the animal, growth factors, and dietary intake. In the present
study, daily concentrations of Ca excreted in the urine appeared low, compared to other
research data (Schryver et al., 1974; Nielsen et al., 1997). Failure to solubilize calcium
carbonate prior to analysis of urinary Ca may have caused concentrations to be low.
Interestingly,visualappraisalofthc graphs fortotalRBAE ofthethirdmetaearpaland
urinary Ca in the stalled horses indicated an inverse relationship. As total RBAES
appearedtodecrease fi'ometod28,subsequentlyincreasethroughd84,andthen
decrease fiomd84tod l40,urinaryCainthestalledhorsesappearedtoincreasethrough
d28,decreasethroughd84, andthenincreasethroughd 140. DecreasingRBAEs
indicated periods of bone resorption, during which Ca was released ficm bone into the
circulation and ultimately excreted in the urine or mces. In contrast, increasing RBAES
indicated a period of bone formation, hence the reduction in Ca excretion through the
urine. Arnaudetal. (1986) demonstratedanincrease inrenalandfecal excretionofCain
healthy subjects immobilized by bed rest for 20 wk.

No treatment difi‘erencesoccmredbetweenpastmedandstalledhorsesinurinaryP
concentrations, except foratendencyforpasturedhorsesto havegreaterurinaryPatd
140. The amount ofP provided by the diet, which was calculated to meet NRC (1989)
recommendations, wasprobablyadequate to meetthemedsofthe horsesinthisstudy.

BothvitaminDandPTleayamajorroleinmaintainingserumCahomeostasis
by enhancing Ca absorption eflicicncy and mobilizing Ca from stores in the bone (Holick,
1996). Parathyroid hormone synthesis and secretion is stimulated by low serum Ca

concentrations, which, in turn, increases tubular reabsorption ofCa in the kidney and

54

 

 

stimulates the production of active 1,25-dihydroxyvitamin D3 (1,25—(OH)2D3) fi'om 25-
hydroxyvitamin D (Vit D) in the kidney. Concentrations ofVit D have been shown to
accurately reflect concentrations of 1,25—(OH)2D3 in the serum (Kumar, 1990). Nornml
serum Ca concentrations are restored by the direct stimulation of osteoclastic resorption
by 1, 25-(01'Dm3 and PTH as they increase differentiation ofpreosteoclasts into new
osteoclasts and increase the activity of existing osteoclasts (Jee, 1988). Additionally, the
efliciency of intestinal Ca absorption is increased by 1, 25-(OH)2D3 (Holick, 1996). Serum
Ca, once withinanornnl range, regulates synthesis and secretionofPTH fi'omthe
parathyroid gland (Pocotte et al., 1991). The role ofVit D in bone metabolism appears to
be the maintenance of adequate mineral concentrations for the deposition of
hydroxyapatite intheboncmatrix, accomplishedbymaintaining intestinalCaabsorption
efficiency (Holick, 1996).

Noueamemdifl'erenccsoccmredinserumVitDandserumPTHconcenu'ations
in the present study. Holick (1994) reported that 80 to 90% ofthe human body's
requirementforvitaminDcanbeobtainedfiomexposureto sunlightalone. Pastured
homeshadfieeaccesstonmlightexceptforshon20m30mmmpefiodsthatwem
conductedinaroundpenor indoorarena. Stalledhorseshadaccesstodirectsunlight
whilebcing walkeddailyonameclmnicalwalkerduringthe 84-dpre-trainingperiodand
onnon-riding daysduringtheS6—dtrainingperiod, aswellasduringtrainingperiods
conducted in an outdoor arena or on the exercise track. Additionally, each horse had
accessto sunlightthroughawindowineachindividualstall. AdequateVitDwouldallow
Ca absorption efliciency, and, in turn, mineral homeostasis to be maintained (Holick,

1996). Adequate serum Ca, demonstrated in the present study, would subsequently

55

 

7
I , r .
. a
o
o
. I 0
. . , . -
1 , I ' . ‘ .
. ‘ 7 -
. . , r

 

 

v .
. 1 " t o v o
a
u . 6
5‘ . . . .
. . . . . .
. I
u . , 7 . '
a u
1 v
. I v , a . .
. . a ' .

suppressPTH synthesisand secretion (PocotteetaL, 1991). RousseletaL (1987)
dennmfiatedthflnnnipuhtionofserumCawmenfiafiommheahhyhorsesproducedm
expectedfluctuationinPTHconcentrations. ResultsofthisstudysupportDalinand
Jefl‘cott(1994)intheirreportthatPTT-IandVichavelimitedusefulnessasmarkersof
bone metabolism, under conditions of adequate sunlightexposure.
Stallingdidnotappeartoinfluencebodyweight,heightatwithers,third
netacmpalcncumferememrbodymndifionscomasmueaummdifi‘emmesoccmred
inanyofthesebodymeasurcments. Bodyweightinbothtreatmentgroupsincreascd
throughthestmtoftrainingatd84,whichisatmhutabletonornnlgrowth. Body
conditionscoredmppedforbothstalledandpasmredhorsesbetweend84andd112and
subsequemlymcreasedmroughdl40,mostlikelyduetomemcreasedemrgymeds
associatedwithtraining. Horseshousedinstallsappearedtobeheavierinweight
mnrparedtopasuuedhomesalmoughbodywndifionscoreswemmtdiflemmbetween
treatments. Thevisualappearanceofweightgaminthestalledhorseswasmostlikelydue
toagreaterhayhuake,comparedtothehorsesnninminedonpastme,causing
enhrgement of the gasuointestinaluactsofthestalledhorses.
Inhialcondifionmgdidmtappearmalleviatethenegafiveefl‘ectsofstaflmgon
boneformation,asdemonstratedbylowerlateralRBAEsinthestalledhorsestlnninthe
pastlnedhorsesduringthetrainingperiod. Perhapsthemagnitudeofstrainappliedtothe
metacarpal bone during training was insufficient to elicit an adaptive response (Lanyon,
1987). Microstrain applied to bone will increase linearly with an increase in velocity, thus
galloping will place more strain on the load-bearing bones than walking or trotting

(Nunamaker, 1986; Pratt, 1982). In the present study, based on observations of elective

56

exercise, each pastured horse cantered an average of 674 strides per day and each stalled
horse cantered only 10 strides per day on the mechanical walker, therefore the pastured
horseshadfieechoiceastomespeedusedwhflespeedwasresuictedmthestalbd
horses. A study of competitive-age horses suggests that introducing speed work, or short,
firstsprints, intotheearly stages ofatrainingprogram, orevenpriortothestartof
training, may place sufficient strain on the bone to prevent bone atrophy and facilitate
normal bone growth (Nlmarmker et al., 1990). Research conducted by Bruin (1993)
involved repeatedly sprinting threemo-old weanlings at short distances over a cement
smfacecoveredwithapproximateltho 3 cmofsandthreetimesperwkfoer mo.
AhhoughnobiochemicalorbonemineralcomemmeasmenrentsweretakenaIZ1/2yrof
age, these horses were competing in 3—day eventing without nruch lameness. In essence,
the most eflicient and effective means of skeletal development appears to be short, fast
sprinting, nottheendmanwtrainingtypicallyenforced intheearlystages ofarace
conditioning program. This modification in training may result in the development of a
skeletal structure appropriately prepared for both training and competition.

A study by Rubin and Lanyon (1985) reported a dose-response relationship
between the magnitude ofstrain externally applied to the ulna ofskeletally mature turkeys
andthebonennsspresent. Strains belowacertainpeakmagnituderesultedinbone loss,
but above that magnitude bone formation exceeded any remodeling activity, resulting in
bone mineral deposition and thus a stronger skeletal structure. Strain rate and distribution
have also been shown to influence the remodeling process (Rubin and Lanyon, 1985).
Caution should be used, however, in implying that the more diverse the training program,

andthemfomflmmomsfiahmagnfiudesmtesmddismbmionsthmmebonehasbeen

57

allowedtoadaptto, thesmallerthepotentialforirrjuryduetothegreaterstrengthofthe
bone. The equine skeletal system must be gradually conditioned for the particular activity
inwhichthehorseisbeingpreparedto compete, inordertomaximizethestrengthand
resistamemlhflmeofbomvfimmebastamoumOfbomfissuenecessaryJedlwmgthe
energynecessaryto mintainandtransportthebonetissuemubin, 1984; Lanyon,1987).
Simply, racehorses should be conditioned with at least sorm speed work, notjust hand-
walking. Themore intensetheactivity,assumingthedurationisnotextensiveandthe
force not too great so as to cause irreparable damage, the greater the adaptive response
and ultimately the skeletal strength and durability (Lanyon, 1989). While Standardbred
horses are conditioned at speeds at which they compete, Thoroughbred horses are often
conditioned at speeds much lower than what they are expected to compete at during a
race. The incidence of dorsal metacarpal disease is much lower in Standardbred horses.
Hence, young Thoroughbred racehorses that are still growing are most likely at a higher
risk of injury due to their inappropriately prepared skeletal structure (Nunannker et al.,
1990). Rubin and Lanyon (1984) also determined that the daily duration of training need
not be excessive in order to stimulate bone. Using a functionally isolated, intact rooster
ulna, artificiallyloadingthebonewith36straincyclesperdayfor6wkresultedina33%
increase in bone mineral content. Beyond 36 cycles per day, no additional new bone
fornmion occurred, indicating that a greater response fiom the cells responsible for the
adaptive mechanism of bone remodeling did not occur. Only 4 peak strain cycles, similar
to fast strides in the horse (Rubin and Lanyon, 1982), per day were determined necessary
to prevent disuse osteoporosis in the master ulna.

Although it was not tested in this study, fi'ee access to exercise may have provided

58

sufliciern loading onthe legs ofpastured horsestopromote normal bone growth. The
studybyNielsenetal. (1997) suggeststhatthehorsesmayhave experiencedgreater
mechanicalloadingontheirlegswhileonpasture,priortod00ftheproject,thanduring
theinitialstagesofconditioning, duringwhichtimetheyweremaintainedinstalls. Results
of the present study indicate that housing yearling/two-yr-old horses in stalls without
accesstoforcedorfi'eeexercisemayimpairnornmlbone growth, comparedwithhorses

maintained on pasture.

59

CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS

A major concern with young performance horses is the high incidence ofskeletal
mjmyassociatedwnhaskektalsuuamemmispoorlypmpmedmhandlethesuessof
training and competition. Young, growing horses are routinely housed in stalls in
preparation for yearling sales or the commencement of training, without any exercise more
intensethanbeinghand-walked. Resultsofthepresentstudysuggestthathousing
yearling/two-yr-old horses in stalls with limited access to exercise may negatively afl’ect
bone growthcomparedtotlntexperiencedbyhorsesallowedto remainonpastlneand
exercise freely.

Lirnitingahorse’saccesstofi'eeexercisecanresultinareductioninstrainapplied
to the bone, stimulating bone mineral resorption as part of the bone’s adaptive response
(Burr et al., 1989). Implications of this phenomena to the industry involve not only
yearlings stalled in preparation for sales or training, but also young performance horses
laidupfollowinginjuryandsubsequentlyreturnedtotraining. Resultsofthepresentstudy
agreewithMoyerandFisher(l991)insuggestingthattrainingshouldresumeatapoint
retroactive to when injury occurred to account for the possible bone loss associated with a
reduction in physical activity, thus preventing reoccurring or additioml injuries.
Bloomfield (1997) reported that after prolonged bed rest inhealthymale subjects, the loss
in bone rmss was not fiilly regained after 6 mo ofnornml weightbearing activity. Even
more significant, lossesinmusclemassandstrengthresultingfi'ombedrestwerefully
reversed weeks or even months before bone losses were recovered, thus contributing to

the risk ofinjury. Weinreb et aL (1997) conducted a study to investigate the recovery of
bone loss during a reloading period following a short period of unloading in young,
growingrats. Afteron1y9 dofthe complete lmloading ofonehindlirnbbyexternal
fixation, complete recovery of cancellous and cortical bone masses occurred after 2 and 3
wk, respectively, indicatingthattherateofbone losswasmuchfirsterthantherateof
bone mass recovery.

While incorporating speed work into a conditioning program is one option for
facilitating normal bone growth, and thus possibly preventing injm'ies that may delay or
eventerminateacompetitive career, other precautionarymeasuresnnybe gleanedfi'om
this study. Allowing young horses primarily confined to stalls, and exercised under
structuredandsupervisedcontrol, accessto freeexercise onpasturemaysupplythe
suficient loading necessaryto strengthenthe skeletal system, aswellasthe opportunity
for young horses to learn coordination and social skills. Fmther research is necessary to
determinewhatlengthoftimeonpastureisnecessaryto establishaskeletalstructurethat
isbetter equipped to handlethe stressoftraining andconrpetition. Providing young,
growing horseswithanenvironmentthatenhancesbone development mayresultina
skeletal structure that is more functionallycapable ofwithstandingthe strains encountered
during training. However, if free exercise is not a viable option or is out of the trainer’s
control, training must be modified accordingly to account for the possible bone loss
experienced as a result of stalling. The younger the horse, the greater the potential for
injury and benefit, therefore training must be monitored carefirlly (Ordidge, 1985).

Urinary deoxypyridinoline and serum osteocalcin appeared to be useful markers of

bone metabolism in the present study. In contrast, the lack oftreatrnent difi‘erences in

61

urinary pyridinoline concentrations suggest that pyridinoline's usefulness as a marker of
bone resorption in limited. Deoxypyridinoline and osteocalcin clearly showed an increase
in bone resorption and a decrease in bone formation, respectively, in the stalled treatment
group shortly after placement in stalls. The pastured horses maintained relatively stable
concentrations of both deoxypyridinoline and osteocalcin, indicating little alteration in
osteoclastic and osteoblastic cell activity, which would be expected during a steady rate of
bone growth. The change in bone resorption and formation activity in the stalled horses
wassupportedbythedecreaseinbonemineralcontent inboththelateralandmedial
cortices ofthethirdmetaearpalinthe stalledgroup. Thus,thenoninvasive radiographic
method of evaluating bone and the biochemical markers of bone resorption and formation
appear to work in concert and suggest the usefulness ofusing both techniques in the

determination of alterations in bone metabolism in a research setting.

62

APPENDD( A

63

 

APPENDIX A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Means Tables
Appendix Table 1A. Means table for body weight (kg). ,

BEL-”W 0 28 J 56 84 112 140
Stall 348.769 367.500 389.432 399.663 392.831 405.060
SEM 10.165 10.580 11.030 11.419 11.296 11.090
Pasture 342.386 358.807 368.864 376.818 377.371 367.213
SEM 10.114 10.580 11.030 11.389 11.396 11.276

Appendix Table 2A. Means table for height at withers (cm).
. Day 0 28 __W 56 84 112 140
Stall 142.625 142.500 142.688 145.688 146.525 145.81 1
SEM 1.045 1.083 1.092 1.388 1.520 1.525
Pasture 143.031 142.813 143.313 144.625 145.433 146.494
SEM 1.045 1.083 1.092 1.388 1.548 1.575
Appendix Table 3A. Means table for third metacarpal circumference (cm).

Day 0 28 “WWWW56 84 112 140
Stall 18.375 18.344 18.438 18.563 18.888 18.625
SEM 0.373 0.302 0.240 0.277 0.626 0.337
Pasture 18.563 18.813 18.625 18.781 19.790 19.040
SEM 0.373 0.302 0.240 0.277 0.651 0.358

Appendix Table 4A. Means table for body condition score. _
WDay 0 2&- 56 84 112 140
Stall 6.219 6.500 6.688 6.625 5.885 6.699
SEM 0.131 0.120 0.112 0.177 0.150 0.175
Pastme 6.125 6.25 6.375 6.500 5.900 6.400
SEM 0.131 0.120 0.112 0.177 0.161 0.190
Appendix Table 5A. Means table for charge in lateral RBAE (nun A41).
Day - 2§__- .. .56 -, 84 _-_-_ A ll}- - 140
Stall -0.928 -0.924 -0.361 -0.721 -0.812
SEM 0.271 0.297 0.478 0.472 0.397
Pasture -0. 124 0.123 0.427 0.542 0.370
SEM 0.271 0.307 0.493 0.505 0.431

Appendix Table 6A. Means table for change in medial RBAE (mm Al).

 

 

 

 

 

 

 

 

 

 

 

Day 28 56 84 112 140
Stall -1.028 -0.868 0.245 -0.478 0.061
SEM 0.269 0.358 0.685 0.530 0.286
Pasture -0.222 0.093 0.377 0.843 0.521
SEM 0.269 0.368 0.705 0.565 0.310
Appendix Table 7A. Means table for serum osteocalcin (ng/ml).

Day 0 14 28 42 56 70
Stall 32.818 28.580 34.231 34.947 29.463 31.842
SEM 2.966 2.569 2.888 2.311 2.761 3.172
Pasture 37.653 35.820 35.663 34.022 33.130 32.189
SEM 2.616 2.266 2.547 2.038 2.435 2.797

Day 77777 84 W ., W WW _ W98 .. W __112- _ _ 126 _W W W140W _W
Stall 34.566 35.205 33.781 30.312 31.961
SEM 2.848 3.740 4.765 3.968 4.727
Pasture 31.115 26.373 39.976 21.188 44.230
SEM 2.512 3.594 4.994 4.277 5.147
Appendix Table 8A Means table for urinary deoxypyridinoline (nM/mM).
Day 0 W W __ 28 W 56W W - _ W 84 _- ., - W112W - 1.4Q-_--_-
Stall 7.541 9.558 7.727 6.791 7.278 7.488
SEM 0.830 0.713 0.561 1.811 0.644 1.082
Pasture 6.738 6.575 6.989 7.936 6.385 7.242
SEM 0.830 0.713 0.561 1.811 0.644 1.082
n = 4 for each treatment group
Appendix Table 9A. Means table for urinary deoxypyridinoline (g/d).
Day 0 28 56 84 112 140
Stall 0.292 0.369 0.253 0.258 0.333 0.341
SEM 0.048 0.037 0.028 0.062 0.033 0.054
Pasture 0.290 0.275 0.275 0.333 0.294 0.352
SEM 0.048 0.037 0.028 0.062 0.033 0.054

 

n = 4 for each treatment group

65

APPENDIX B

APPENDIX B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Proc Mixed Tables
Appendix Table 1B. Proc mixed table for body weight (kg).

- Source _-____-_-_ NDF .. _ _W DDF F-Value P-Value
Treatment 1 14 1.57 0.2307
Day 5 58 30.70 0.0001
Day‘Treatment 5 58 6.37 0.0001
Appendix Table 2B. Proc mixed table for height at withers (cm).

Source NDF DDF F-Value P-Value
Treatment 1 14 0.00 0.9899
Day 5 60 7.07 0.0001
Day‘Treatment 5 60 1.50 0.2045
Appendix Table 3B. Proc mixed table for third metacarpal circumference (cm).
Source ,_ NDF _W . DDF F-Value P-Value
Treatment 1 14 0.88 0.3634
Day 5 60 1.09 0.3779
Day‘Treatment 5 60 0.64 0.6678
Appendix Table 4B. Proc mixed table for body condition score.

_ _Source NDF DDF F-Value P-Value __-
Treatment 1 14 1.80 0.2014
Day 5 60 12.71 0.0001
Day‘Treatment 5 60 0.79 0.5609
Appendix Table 5B. Proc mixed table for lateral RBAE (mm Al).

- Source NDF DDF F-Value WW _. P-Value -
Treatment 1 14 0.22 0.6446
Day 5 58 1.74 0.1404
Day‘Treatment 5 58 1.15 0.3465
Appendix Table 6B. Proc mixed table for medial RBAE (mm Al).

Source NDF DDF F -Va1ue P-Value
Treatment 1 14 0.09 0.7716
Day 5 58 4.90 0.0008

Day‘Treatment 5. _ 58 _ 2.13 0.0741

 

67

Appendix Table 7B. Proc mixed table for change in lateral RBAE (mm Al).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Source NDF DDF F-Value P-Value
Treatment 1 14 6.35 0.0245
Day 4 44 0.81 0.5260
Day*Treatment 4 44 0.29 0.8849
Appendix Table 8B. Proc mixed table for change in medial RBAE (mm Al).
Source? NDF , w _ _ _ . -I?DF w. , . - -F-Yalue. , _P-VaI-A- _A
Treatment 1 14 2.49 0.1371
Day 4 44 3 .65 0.01 18
Day*Treatment 4 44 1 .28 0.2916
Appendix Table 9B. Proc mixed table for total RBAE (mm2 Al).

Source NDF DDF F -Value P-Value
Treatment 1 14 0.02 0.8996
Day 5 58 2.67 0.0308
Day*Treatment 5 58 1 .26 0.2914
Appendix Table 10B. Proc mixed table for serum osteocalcin (ng/ml).

Source NDF DDF F—Value P-Value
Treatment 1 14 0.23 0.6371
Day 10 121 4.50 0.0001
Day*Treatment 10 121 3.66 0.0003
Appendix Table 11B. Proc mixed table for urinary deoxypyridinoline (nM/mM).
Source NDF DDF F -Value P-Value
Treatment 1 6 1.12 0.3301
Day 5 30 0.83 0.5373
Day*Treatment 5 30 1.18 0.3398

 

n = 4 for each treatment group

Appendix Table 12B. Proc mixed table for 11an deoxypyridinoline (g/d).

 

 

Source NDF DDF F -Value P-Value
Treatment 1 6 0.01 0.9232
Day 5 30 2.42 0.0590
Day*Treatment 5 30 2.33 0.0667

 

r1 = 4 for each treatment group

68

Appendix Table 13B. Proc mixed table for urinary pyridinoline (nM/mM).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-Sg_meA-ANDF-__--_----_--1291:__ _____-__F_-__Y8!uc P-Value
Treatment 1 6 2.22 0.1867
Day 5 29 2.92 0.0297
Day‘Treatment 5 29 0.87 0.5159
n = 4 for each treatment group
Appendix Table 143. Proc mixed table for serum calcium (mg/d1).

. Source NDF __ _ DDF F-Value P-Value
Treatment 1 14 0.08 0.7850
Day 10 121 21.00 0.0001
Day‘Treatment 10 121 1.87 0.0552
Appendix Table 15B. Proc mixed table for serum phosphorus (mg/d1).

Source NDF DDF F-Value P-Value
Treatment 1 14 6.97 0.0194
Day 10 121 16.20 0.0001
Day‘Treatment 10 121 6.00 0.0001
Appendix Table 163. Proc mixed table for urinary calcium (g/d).

Source NDF DDF F-Value P-Value WWWWW
Treatment 1 6 2.75 0.1486
Day 5 30 7.10 0.0002
Day‘Treatment 5 30 0.40 0.8432
n = 4 for each treatment group
Appendix Table 17B. Proc mixed table for urinary phosphorus ngld).

Ame - NDF A -ADQF F-Value P-Vahle
Treatment 1 6 0.81 0.4040
Day 5 30 3.03 0.0249
Day‘Treatment 5 30 3.67 0.0104

 

n=4foreachtreatmentgroup

69

LITERATURE CITED

70

LITERATURE CITED

Amud, S.B., V.S. Schneider, and E. Morey-Holton. 1986. Efi'ects of mactrvrty on bone
and calciurnmetabolism. In: H. SandlerandJ. Vemikos (Eds) Inactivity:
Physiological Efl‘ects. p. 49-75. Academic Press, Inc., Orlando, FL.

thk, A., S.L. Rabton, S.A. Shapses, L. Suslak-Brown, P.A. Schoknecht. 1997.
Skeletal growth patterns in Standardbred foals frombirthto 1 year. Proc.15"I
Equine Nutr. Phys. Symp. p. 326-327.

Bloomfield, SA. 1997. Changes in musculoskcletal structm'e and function with
prolonged bed rest. Med. Sci. Sports Ex. 29(2):197-206.

Bronner, F. 1993. Nutrient bioavailability, with special reference to calcium. J. Nutr.
123:797-802.

Brown, J.P., L. Malaval, MC. Chapuy, P.D. Delrms, C. Edouard and P.J. Meunier.
1984. Serum bone gla-protein: a specific marker for bone formation in
postmenopausal osteoporosis. Lancet. 1:1091-1093.

Bruin, G. 1993. The effect of exercise on the incidence of osteochondrosis in young
horses. J. Equine Vet. Sci. 13(3):134-135.

Buckingham, S.H.W., and LB. Jefl‘cott. 1991. Skeletal effects of a long term
submaxinnl exercise programme on Standardbred yearlings. In: J.R. Gillespie and
NE. Robinson (Ed.) Equine Exercise Physiology 3. p. 411-418. ICEEP
Publications, Davis, CA.

Burr, D.B., M.B. Schamer, K.H. Yang, D.D. Wu, M. Lukoschek, D. Kandzari, N.
Sivaneri, J.D. Blaha and EL. Radin. 1989. The effects of altered strain
enviromnents on bone tissue kinetics. Bone. 10:215-221.

Carter, S.D., G.L. Cromwell, T.R. Combs, G. Colombo and P. Fanti. 1996. The
determinationofsemmconcenfiationsofosteocalciningrowingpigsandits
relationship to end-measures of bone mineralimtion. J. Anirn. Sci. 74:2719-2729.

Currie, B.W. 1988. Growth and development. In: Structure and chtion of Domestic
Animals. p. 293-313. Butterworth Publishers, Stoneham, MA.

Cymbaluk, N.F., and 6.1. Christison. 1989. Effects of dietary energy and phosphorus

content on blood chemistry and development of growing horses. J. Anim. Sci.
67(4):951-958.

71

Dalin, G., and LB. Jefl‘cott. 1994. Biomechanics, gait, and conformation. In: D.R.
Hodgson and R.J. Rose (editors) The Athletic Horse: Principles and Practice of
Equine Sports Medicine (1it Ed). p. 2748. WE. Saunders Company,
Philadelphia, PA.

Delrnas, PD. 1993. Biochemical markers of bone tin-hover. J. Bone Min. Res.
8(2):SS49-S555.

Duncan, JR, and K.W. Prasse. 1986. Veterinary Laboratory Medicine. Iowa State
University Press, Ames, IA.

El Shorafa, W.M., J.P. Feaster and EA. Ott. 1979. Horse metacarpal bone: age, ash
content, cortical area and firilure stress interrehtionships. J. Anim. Sci. 49(4):979-
982.

Evans, J.W., A. Berton, H. Hintz, and LD. Van Vleck. 1990. Anatomy, physiology, and
care ofthe feet and legs. In: The Horse (2ml Ed). p. 683-752. W.H. Freeman
and Company, New York, NY.

Fleming, R11, C.C. Whitehead, D. Alvey, N.G. Gregory and LJ. Wilkins. 1994. Bone
suuchueandbreakingsuengthinlayinghenshousedindifieremhusbmdry
system Brit. Poult. Sci. 35:651-662.

Frey, K.S., G.D. Potter, T.W. Odom, D.M. Senor, V.D. Reagan, V.H. Weir, J. Elslander,
SP. Webb, E.L. Morris, WB Smith, andKE Weigand. 1992. Plasrnasilicon
andradiographic bone densityinweanlingquarterhorsesfedsodimnzeoliteA.
Proc.12“' Equine Nutr. Phys. Symp. p. 291-295.

Frost, HM. 1987. Bone “mass” and the “mechanostatz” a proposal. Amt. Rec. 219:1-
9.

Gomez, Jr., B., S. Ardakani, 3.]. Evans, L.D. Merrell, D.K. Jenkim, and V.T. Kung.

1996. Monoclonal antibody assay forfi'ee minarypyridiniumcross-links. Clin.
Chem. 42(8):1168-1175.

Gomez,B.,CA.BallyDK.Jenkms,R.JKe1rn,Jr.andS.Seyedin.1994.Anenzyme
irmmmoassayforintact,newlysynthes1zedosteocalcm. amarkerforbone
fornntion. Int. Conf. Prog. BoneMin.Res. p. 533-540.

Henneke, D.R., G.D. Potter, and J.L. Kreider. 1984. Body condition during pregnancy
and lactation and reproductive efficiency of mares. Theriogenology. 21(6):897-
909.

Holick, MF. 1994. Vitamin D: new horizons for the 21st century. Am. J. Clin. Nutr.
60:619-630.

Holick, MF. 1996. VitaminDandbonehealth. J. Nutr. 126:1159S-ll648.

Jaworski, Z.F.G. 1984. Lamellar bone turnover system and its effector organ. Calcif.
Tissue Int. 36:S46-855.

Jee, W.S.S. 1988. The skeletal tissues. In: L. Weiss (Ed) Histology: Cell and Tissue
Biology (5" Ed). p. 207-254. Elsevier Biomedical, New York, NY.

Jefi‘cott, L.B., S.H.W. Buckingham, and RN. McCartney. 1987. Noninvasive
measurementofbone qualityinhorsesandclmnges associatedwithexercise. In:
J.R. Gillespie and NE. Robinson (Ed) Equine Exercise Physiology 2. p. 615-
630. ICEEP Publications, Davis, CA.

Kannus, P., L. Jozsa, M. Kvist, T.L.N. Jarvinen, V. Maunu, T. mama, and M. Jarvinen.
1996. Expression of osteocalcin in the patella of experirmntally immobilized and
remobilized rats. J. Bone Min. Res. 11(1):79-87.

Kirmnel, DB. 1993. A paradigm for skeletal strength homeostasis. J. Bone Min. Res.
8(2):SS 15-8522.

Knowles, T.G., and D.M. Broom. 1990. Limb bone strength and movement in laying
hens fi'om difl‘erent housing systems. Vet. Rec. 126:354-356.

Krook, L. 1968. Dietary calcium-phosphorus and lameness in the horse. Cornell Vet.
58(Suppl. 1):59-73.

Krook, L., and J.E. Lowe. 1964. Nutritional secondary hyperparathyroidism in the horse.
Pathol. Vet. 1(1):1-98.

Kunnr, R. 1990. VitaminD metabolismandmechanisrm ofealciumtransport. J. Am.
Soc. Nephrol. 1:30-42.

Lanyon, LE. 1984. Functional strain as a determinant for bone remodeling. Calcif.
Tissue Int. 36:S56-S6l.

Lanyon, LE. 1987. Flmctional strain in bone tissue as an objective, and controlling
stimulus for adaptive bone remodelling. J. Biomech. 20(11/12):1083-1093.

Lanyon, LE. 1989. The equine skeleton. In: W.E. Jones (Ed) Equine Sports
Medicine. P. 53-58. Lea & Febiger, Philadelph'm, PA.

Lawrence, L.A., E.A. Ott, G.J. Miller, P.W. Poulos, G. Piotrowski and R.L. Asquith.

1994. Mechanical properties ofequine third metacarpals as affected by age. J.
Anim. Sci. 72:2617-2623.

73

LeBlanc, A.D., V.S. Schneider, H.J. Evans, D.A. Engelbretson, and JM. Krebs. 1990.
Boneminerallossandreeoveryafter 17weeksofbedrest. J. BoneMin.Res.
5(8):843-850.

Lepage, O.M., L. DesCoteaux, M. Mareoux and A. Tremblay. 1991. Circadian rhythms
of osteocalcin in equine serum. Correlation with alkaline phosphatase, calcium,
phosphate and total protein levels. Can. J. Vet. Res. 55:5-10

Lepage, D.M., M. Marcoux, and A. Tremblay. 1990. Serum osteocalcin or bone gla-
protein,abiochemicalmarkerforbone metabolisminhorses: difi‘erencesinserum
levels with age. Can. J. Vet. Res. 54:223-226.

Maenpaa, P.H., A. Pirskanen, and E. Koskinen. 1988. Biochemical indicators of bone
formationinfoalsaflertransferfi'ompastmeto stablesforthewintermontlm. Am.
J. Vet. Res. 49(11):]990-1992.

Marchant, J.N., and D.M. Broom. 1996. Efi’ects of dry sow housing conditions on
muscle weight and bone strength. Anim. Sci. 62:105-113.

Marks, SC, and SN. Popofl'. 1988. Bone cell biology: the regulation of development,
structure, and fimction in the skeleton. Am. J. Anat. 183:1-44.

McCarthy, R.N., and LB. Jefl‘cott. 1992. Efi'ects of treadmill exercise on cortical bone in
the third metacarpus of young horses. Res. Vet. Sci. 52:28-37.

McCartney, R.N., and LB. Jeffcott. 1987. Combined 2.25 MHz ultrasound velocity and
bonemineraldensitymeasurementsintheequine metacarpusandtheirinvivo
applications. Med. Biol. Eng. Comp. 25:620.

Meakim, D.W., E.A. Ott, R. L. Asquith, and J.P. Feaster. 1981. Estimation of mineral
content of the equine third metacarpal by radiographic photometry. J. Anim. Sci.
53(4):]019-1026.

Moyer, W., and J.R.S. Fisher. 1991. Bucked shins: efl‘ects of (1ng track surfaces and
proposed training regimens. Proc. Am. Assoc. Equine Pract. 36:541-547.

Nielsen, B.D., and GD. Potter. 1997. Accounting for volumetric difi‘erences in estinntes
of bone mineral content from radiographic densitometry. Proc. 15th Equine Nutr.
Phys. Symp. p. 367-369.

Nielsen, B.D., G.D. Potter, L.W. Greene, E.L. Morris, M. Mm'ray-Gerzik, W.B. Smith,
and M.T. Martin. 1998. Characterization of changes rehted to mineral balance
and bone metabolism in the ymmg racing Quarter Horse. J. Equine Vet. Sci.
18(3):]90-200.

74

Nielsen, B.D., G.D. Potter, E.L. Morris, T.W. Odom, D.M. Senor, J.A. Reynolds, W.B.
Smith, andM.T. Martin. 1997. Chmgesinthethirdmetacarpalboneand

fi‘equencyofboneinjmiesinyomgQuarterHorsesduringracetraining-
observations and theoretical considerations. J. Equine Vet. Sci. 17(10):541-549.

Nishimoto, S.K., and RA. Price. 1980. Secretion of the vitamin K-dependent protein of
bone by rat osteosarcoma cells. Evidence for an intracellular precursor. J. Biol.
Chem. 255:6579-6583.

Norgaard-Nielsen, G. 1990. Bone strength of laying hens kept in an alternative system,
compared with hens in cages and on deep-litter. Brit. Poult. Sci. 31:81-89.

Norwood, G.L. 1978. The bucked-shin complex in Thoroughbreds. Proc. 24Ill Am.
Assoc. Equine Pract. p. 319-336.

Nunamaker, D.M. 1986. The bucked shin complex. Proc. 32"l Am. Assoc. Equine Pract.
p.457-460.

Nunamaker, D.M., D.M. Butterweck, and M.T. Provost. 1989. Some geometric

properties ofthethirdmetaearpalbone: acomparisonbetweentheThoroughbred
and Standardbred racehorse. J. Biomech. 22(2):]29-134.

Nunamaker, D.M., D.M. Butterweck, and M.T. Provost. 1990. Fatigue fixtures in
Thoroughbred racehorses: relationships with age, peak bone strain, and training. J.
Orthop. Res. 8:604-611.

NRC. 1989. Nutrient Requirements of Horses. (5th Ed.). Natioml Academy Press,
Washington, DC.

Ordidge, RM. 1985. The equine metacarpus. Part 2: the cannon bone. Vet. Annual.
25:192-200.

Pocotte, S.L., G. Ehrenstein, and LA. Fitzpatrick. 1991. Regulation of parathyroid
hormone secretion. Endocr. Rev. 12:291-301.

Porr, C.A., E.A. Ott, E.L. Johnson, and J.B. Madison. 1997. Bone mineral in young
Thoroughbred horses is afl‘ected by training. Equine Pract. 19(8):28-31.

Pratt, G.W. 1982. Theresponse ofhighlystressedboneintheracehorse. Proc. 28'”Am.
Assoc. Equine Pract. p. 31-37.

Price, P. 1982. Osteocalcin. In: W.A. Peck (Ed) Bone and Mineral Research, Volume
1. p.157-190. Excerpta Medica, Ansterdam.

75

 

Price, 1.8., B. Jackson, R. Eastell, A.M. Wilson, R.G.G. Russell, L.E. Lanyon and AE.
Goodship. 1995. The response of the skeleton to physical training: a biochemical
study in horses. Bone. 17(3):221-227.

Price, P.A., J.G. Parthemore, L.J. Defios and SK Nishimoto. 1980. New biochemical
rmrkerforbonemetabolism: measurement byradioimmunoassayofbonegla
proteinintheplasmaofnormalsubjectsandpaticntswithbonedisease. J. Clin.
Invest. 66:878-883.

Price, P.A., J.W. Poser, and N. Raman. 1976. Primary structure ofthe gamma-
carboxyglutamic acid-containing protein fiom bovine bone. Proc. Nat. Academy
Sci. 73:3374-3375.

Raisz, L.G., B. Wiita, A. Artis, A. Bowen, S. Schwartz. M. Trahiotis, K. Shoukri, and J.
Smith. 1996. Comparison of the efl‘ects of estrogen alone and estrogen plus
androgen on biochemical mkers of bone formation and resorption in
postmenopausal women. J. Clin. Endo. Metab. 81(1):37-43.

Riggs, CM, and GP. Evans. 1990. The microstructural basis of the meclnnical
properties of equine bone. Equine Vet. Ed. 2(4):]97-205.

Rivera, B., S. Benjamin, B.D. Nielsen, J.E. Shelle and A.J. Zanella. 1997. Behavioral
andphysiological responses ofhorsesto initialtraining, thecomparisonbetween
pasture versus stalled horses. Int. Soc. Anthrozoology Conf. July 24-25. Boston,
MA.

Robins, S.P., H. Woitge, R. Hesley, J. In, S. Seyedin and MJ. Seibel. 1994. Direct,
enzyme-linked immunoassay for urinary deoxypyridinoline as a specific marker
for measuring bone resorption. J. Bone Min. Res. 9(10):1643-1649.

Roussel, A.J., Y.C. Lin, J.R. Strait, and RD. Modramky. 1987. Radioimmunoassay for
parathyroid hormone in equids. Am. J. Vet. Res. 48(4):586—589.

Rubin, CT. 1984. Skeletal strain and the functional significance of bone architecture.
Calcif. Tissue Int. 36:811-818.

Rubin, C.T., andL.E. Lanyon. 1982. Limb mechanicsasafunction of speedand gait: a
studyoffimctionalstrainsintheradiusandtibiaofhorseanddog. J. Exp. Biol.
101:187-211.

Rubin, C.T., and LE. Lanyon. 1984. Regulation of bone fonmtion by applied dynamic
loads. J. Bone Joint Surg. 66A(3):397-402.

Rub'm, C.T., and LE. Lanyon. 1985. Regulation of bone mass by mechanical strain
magnitude. Calcif. Tissue Int. 37:411-417.

76

SAS. 1997. SAS System for Mixed Models. SAS Institute Inc., Cary, NC.

Schryver, H.F., H.F. Hintz, and J.E. Lowe. 1974. Calcium and phosphorus in the
nutrition of the horse. Cornell Vet. 64:493-515.

Vico, L., D. Chappard, C. Alexandre, S. Palle, P. Minaire, G. Rifl‘at, B. Morukov, and S.
Rakhnnnov. 1987. Efl‘ects ofa 120 day period ofbed-rest on bone mass and
bone cell activities in man: attempts at countermeasure. Bone Min. 2:383-394.

Weinreb, M., H. Patael, O. Preisler, and S. Ben-Shemen. 1997. Short-term healing
kinetics of cortical and cancellous bone osteopenia induced by Imloading during
the reloading period in young rats. Virchows Arch. 431:449-452.

Weryha, G., and J. Leclere. 1995. Paracrine regulation of bone remodeling. Hormone
Research 43:69-75.

Williams, S.N., L.R. McDowell, L.A. Lawrence, N.S. Wilkinson, P.W. Ferguson and
A.C. Wamick. 1991. Criteria to evaluate bone mineralization in cattle: II.
Noninvasive techniques. J. Anim. Sci. 69:1243-1254.

Woo, S.L.Y., S.C. Kuei, D. Amiel, M.A. Gomez, W.C. Hayes, RC. White, and W.H.
Akeson. 1981. The effect ofprolonged physical training on the properties oflong
bone: a study of Wolfi’s law. J. Bone Joint Surg. 63A(5):780-787.

Yeh, J.K., C.C. Liu, and IF. Aloia. 1993. Effects of exercise and immobilimtion on

bone formation and resorption in young rats. Am. J. Physiol. 264(27):E182-
E189.

77

nrcurcaN STATE UNIV. LIBRARIES
l“WIWIIWIWIIIIHIIIIIIHIHHIIHIIIIHHIIHII
31293017141247