CLINICAL ORTHOPAEDICS AND RELATED RESEARCH
Number 3555, pp S105-SI15
Strain Rate and Timing of
Stimulation in Mechanical Modulation
of Fracture Healing
Allen E. Goodship, BVSc, PhD, MRCVS*j**;
James L. Cunningham, PhD, CEng, MI Mech E f;
and John Kenwright, MD f f
Fracture of the long bones results in a repair
process that has the potential to restore the
anatomic morphology and mechanical in-
tegrity of the bone without scar tissue. The re-
pair process can occur in two patterns. In the
first, under conditions of rigid stabilization, di-
rect osteonal remodeling of the fracture line
can occur with little or no external callus, a
process known as direct bone repair. The sec-
ond pattern of repair involves bridging of the
fragments with external callus and formation
of bone in the fracture site by endochondral
healing. This type of repair is known as indi-
rect bone healing and occurs under less rigid
interfragmentary stabilization. The rate of
healing and the extent of callus in this type of
repair can be modulated by the mechanical
conditions at the fracture site. Applying cyclic
interfragmentary micromotion for short peri-
ods has been shown to influence the repair
From the *Royal Veterinary College, University of
London, London, England; **Institute of Orthopaedics,
University College of London, London, England; tDe-
partment of Orthopaedic Surgery, University of Bristol,
Bristol, England; and TtNuffield Department of Or-
thopaedic Surgery, University of Oxford, Oxford, Eng-
land.
Reprint requests to Allen E. Goodship, BVSc, PhD,
MRCVS, Royal Veterinary College, Hawkshead Lane,
North Mymms, Herts, AL9 7TA, UK.
S105
0 1998 Lippincotl Williams & Wilkins
process significantly, and characteristics of
this stimulus influence the healing response
observed. In the current study, a short term in-
terfragmentary cyclic micromovement applied
at a high strain rate induced a greater amount
of periosteal callus than the same stimulus ap-
plied at a low strain rate. This high strain rate
stimulus applied later in the healing period
significantly inhibited the progress of healing.
The beneficial effect of this particular bio-
physic stimulus early in the healing period
may be related to the viscoelastic nature of the
differentiating connective tissues in the early
endochondral callus. In the early endochon-
dral callus, high rates of movement induce a
greater deformation of the fracture fragments
because of the stiffening of the callus. Alterna-
tively, the transduction pathway may involve
streaming potentials as a result of the high
movement rate.
Intact bone responds to short periods of
cyclic mechanical deformation to effect
adaptive changes that optimize the mass and
architecture of the bone to accommodate the
new imposed loading regimen.13 In a similar
manner, the process of bone repair also is
modulated by mechanical infl~ences.3~8-12~~~
There are two basic patterns of repair. The
first is seen in conjunction with rigid stabi-

S106 Goodship et al
lization of the fracture fragments and in-
volves direct osteonal remodeling across the
fracture site at points of contact. Woven bone
forms in the gaps where there is no apposi-
tion of bone fragments. In this type of heal-
ing, there is little or no periosteal callus.'*
This process occurs over a long period and
involves intracortical bone loss, which is a
short term phenomenon.19
The second pattern of healing is known as
indirect repair, which occurs when the frac-
ture fragments are stabilized under less rigid
conditions and involves two separate
processes of ossification. The first process is
intramembranous ossification, in which fi-
brous and osteogenic periosteal cells prolif-
erate to form a bulging fibrous bridging cal-
lus. The fibrous tissue then undergoes
progressive intramembranous ossification to
form a bony bridge between the fracture
fragments. In the fracture site, damage to the
blood vessels leads to a fracture hematoma,
which then undergoes differentiation
through granulation tissue, fibrous tissue,
and fibrocartilage to hyaline cartilage, and
then by endochondral ossification to woven
and lamellar bone to give osseous union be-
tween the fracture fragments. The external
callus then is remodeled as a function of
time, and increased mechanical use restores
the mass and architecture of the bone to near
normal values.
In indirect bone repair, the extent and dis-
tribution of periosteal callus together with
the rate of differentiation are influenced by
mechanical conditions applied to the fracture
site during the healing period.3,8-12,24A previ-
ous study showed that short periods of ap-
plied cyclic micromovement on a daily basis
induced an earlier formation of periosteal
callus and increased the rate of healing com-
pared with unstimulated controls. Subse-
quent data8,9J1showed that specific charac-
teristics of the mechanical stimulus, within
the short period over which it is applied, can
themselves have significant effects on the
healing process. Thus too great an initial
magnitude of displacement or the use of too
Clinical Orthopaedics
and Related Research
high an initial force imposed a significant in-
hibition on the healing process.
The preceding data imply that exceeding
the ultimate strain or yield point of tissues in
the early heterogeneous callus mass could
induce additional damage and repair cycles.
Such a mechanism may be present in certain
types of dynamic fixation in which the fixa-
tion device is rigid during the first few
weeks, then allowed to collapse once early
callus is formed.5.16 Under such conditions,
the fracture site is compressed and the gap
size reduced.L.6 However, this is not the same
mechanical stimulation as exists in the appli-
cation of cyclic micromovement maintaining
the original gap size. A reduction in the size
of the fracture gap will enhance the rate of
healing.2 It therefore is important to control
variables such as gap size when trying to elu-
cidate the mechanisms involved in transduc-
tion of mechanical modulation of bone re-
pair.
Previous studies also have indicated that
the repair process is retarded under condi-
tions of insufficient mechanical stimula-
tion.l0J1 Thus the goal is to provide the most
appropriate signal as a function of time to
drive the ossification processes of indirect
bone repair at an optimal rate by removing
the inhibitory influences of the fixation de-
vice and the reduced functional use of the
limb.
As the repair process progresses, there is a
general increase in the stiffness of the differ-
entiating tissues and in stability of the frac-
ture fragments. The material properties of
the early differentiating tissues in indirect
fracture repair can be described as viscoelas-
tic, so the rate of deformation may result in
different levels of deformation. Hence high
rates of deformation will induce a greater
stiffness in these tissues and thus give a
lower displacement for a given load. At these
high rates, the increased callus stiffness will
result in a higher strain on the fragment ends,
and also may induce higher streaming poten-
tials in parts of the callus.I5 Hence the over-
all mechanical conditions in the early soft
Number 3558
October, 1998
callus will be influenced by the rate of defor-
mation.
It also has been shown that strain rate is
an important component of osteogenic me-
chanical stimulation in intact bone.17 In-
creasing the stiffness of the soft callus may
increase the osseous deformation of the hard
callus and adjacent bone fragments, which
should provide a greater overall osteogenic
stimulus. The characteristics of other puta-
tive transduction mechanisms, such as strain
related streaming potentials generated as a
result of charged ions moving in extracellu-
lar fluid, also are influenced by strain rate.
Thus although the variables of displacement
magnitude and applied force have been ex-
amined, the question of what effect displace-
ment rate has on this defined model has not
yet been addressed.
In this study, it was hypothesized that an
increased rate of deformation will enhance
the osteogenic effects of short periods of
cyclic micromovement. It also is postulated
that the maximal effect of such a stimulus
will occur in the early phase of the repair
process. This study was designed to test
these two hypotheses using the same ovine
model used in the authors’ previous studies
on the effects of applied micromovement on
indirect bone
STUDY DESIGN
This study design addressed two hypothe-
ses: (1) the effect of short periods of cyclic
mechanical stimulation on a healing os-
teotomy can be modified by subtle differ-
ences in the rate of interfragmentary cyclic
stimulation; and (2) this procedure is more
effective in stimulating healing if applied in
the early phase of the repair process. In the
first part of the study, two different rates of
interfragmentary movement were compared
with the rate used in earlier studies to evalu-
ate effects of displacement and load. The
highest displacement rate regimen then was
used in the second part of the study to deter-
mine the effect of this mechanical stimula-
Mechanical Modulation ot Fracture Healing S107
tion in relation to when it is applied to the
healing process.
MATERIALS AND METHODS
Model
The previously described9 standardized ovine
model was used. In this model, a middiaphyseal
osteotomy is distracted to form a 3-mm gap and
stabilized with a unilateral external fixator (Ox-
ford MK 11, Johnson and Johnson Orthopaedics,
New Milton, Hants, UK). The fixator was applied
to the cranial aspect of the tibia, using jigs to en-
sure repeatable positioning in relation to defined
anatomic landmarks. The frame geometry also
was standardized using a system of jigs. All surgi-
cal procedures were performed under aseptic con-
ditions that complied with the appropriate
regulatory requirements.
The fixator used six 6-mm Schanz screws,
three inserted on each side of the osteotomy. The
tibia was drilled and pretapped to provide some
radial prestress. Screw sites were dressed with
povidine iodine on a daily basis. The sheep
walked within 24 hours of surgery, and the fixator
frames were left in situ for 12 weeks.
In the first part of the study, three groups of
skeletally mature female sheep were used in
which displacement rates of 2 mm, 40 mm, and
400 mm per second, were applied using a micro-
processor controlled servohydraulic actuator. An
initial displacement of 1 mm was used, applied
with a force of 200 N for 500 cycles at 0.5 Hz for
5 consecutive days per week for 12 weeks. The
stimulus was applied from 1 week postopera-
tively in all three groups.
The second part of the study used an addi-
tional group of six skeletally mature females. The
osteotomies in this group were subjected to the
same stimulus (400 mdsecond of applied cyclic
micromovement) as the preceding groups, but the
stimulus was initiated at 6 weeks postoperatively
when periosteal bridging had commenced.
Assessment of Healing
In vivo assessments were undertaken postopera-
tively and every 2 weeks thereafter to the 12-
week allocated end point.
Radiography
The fixator was mounted in a custom made jig to
ensure standard radiographic views. A lateral

S108 Goodship et al
view radiograph was used because a craniocaudal
exposure was precluded by the presence of the
metallic fixator bar. The xray beam was centered
on the midosteotomy gap. A step wedge was
included on each exposure. Radiographs were as-
sessed subjectively on a blind basis by indepen-
dent orthopaedic surgeons.
Bone Densitometiy
A jig system was used to obtain standard position
dual photon absorptiometry scans (using photon
sources of Iodine 125 and Americium 241). An
initial longitudinal scan was made to ensure loca-
tion of the osteotomy gap after which transverse
scans were made at the midgap level and at a set
distance proximal and distal to the osteotomy
site.
Fracture Stiffness Index
Simultaneous data were collected from a strain
gauge transducer attached to the fixator bar7 and a
force plate during normal walking. Ten records
were used to calculate a mean index of fracture
stiffness? during walking (defined as walking
stiffness index) at each time point. Group means
of these data were plotted as a function of postop-
erative time.
Strain rate = 2 rnrnlsecond Strain rate = 40 rnrnlsecond
Fig 1. Standard view radiographs 12 weeks after surgery show the difference in callus distribution
associated with the three different stimulation protocols. From left to right: strain rate = 2, 40, 400,
and 400 mm per second (late start).
Clinical Orthopaedics
and Related Research
Postmortem Assessments
Mechanical testing was carried out to determine
the torsional stiffness and ultimate torsional
strength of the healing osteotomy and the con-
tralateral intact tibia. The bones were mounted to
give a standard gauge length by embedding the
tibia and the screw holes on each side of the os-
teotomy. Inward rotation was used to simulate in
vivo torsional strain distribution.
Statistical Analysis
The results were analyzed using the Mann-Whit-
ney U test (comparison of postmortem stiffness
and strength) and analysis of covariance (AN-
COVA) using the general linear model (bone min-
eral content and walking stiffness index versus
time). The analysis was done using Minitab
(Minitab Inc, State College, PA) and SPSS for
Windows (SPSS Inc, Chicago, IL).
RESULTS
Radiography
There was an evident difference between
groups with respect to the rate of applied mi-
cromovement (Fig l). A greater formation of
periosteal callus formation was seen in the
Strain rate = 400 mrnlsecond Strain rate = 400 rnmlsecond
(late start)
Number 3558
October, 1998
high displacement rate group than in the in-
termediate or low displacement rate groups.
However, there was a greater degree of con-
solidation and radiodensity of the osteotomy
gap in the intermediate (40 mdsecond)
group. The group subjected to the delayed ap-
plication of the 400 mm per second stimulus
showed a reduced amount and consolidation
of callus at the 12-week end point (Fig 1).
Bone Densitometry
A highly significant (p < 0.005) difference
was observed in the groups as assessed by
analysis of variance (ANOVA) using the
general linear model, with the rate of bone
mineral content increase in the midgap great-
est in the 40 mm per second group (Fig 2).
For the delayed stimulation group, the rate of
bone mineral content increase was lower
than for the similar group, in which the stim-
ulus had been applied from the postoperative
period (Fig 3). The difference between these
groups was highly significant (p < 0.001).
Walking Stiffness Index
There was a highly significant (p < 0.001)
difference between the walking stiffness ob-
tained over time for the three groups. The
rate of increase in walking stiffness index
Bone Mineral Content (glcm)
1000
800
600
400
200
Fig 2. Osteotomy gap bone
mineral content with time after
surgery showing the influence 0
of the applied micromotion rate
(& standard error of the mean, 1 2
n = 6).
Mechanical Modulation of Fracture Healing s109
was greatest in the high displacement rate
group (400 mdsecond), which showed a
continued rate of increase up to the 12-week
end point without a plateau. The 40 mm per
second group showed the next greatest rate
of increase in walking stiffness index, and
the 2 mm per second group the lowest rate of
increase in walking stiffness index (Fig 4).
The delayed stimulation group showed a sig-
nificantly reduced rate of increase in walking
stiffness index compared with the similar
group that had been stimulated throughout
(Fig 5). The difference between the two
groups was highly significant (p < 0.005).
Postmortem Testing
In terms of torsional stiffness and strength,
significantly higher values were seen in the
40 mm per second group (Fig 6) than in the
400 mm and 2 mm per second groups. Sig-
nificant difference was observed in tor-
sional stiffness and strength between the
400 mm and 2 mm per second groups. The
400 mm per second group subjected to the
delayed application of the strain showed a
lower torsional stiffness (not significant, p =
0.17) and strength (significant, p = 0.03)
than the similar group stimulated throughout
(Fig 7).
3 4 5 6 7 8 9 1 0 1 1 1 2
Weeks After Operation
 Clinical Orthopaedics
S110 Goodship et al and Related Research

Bone Mineral Contant (g/cm)

-
600 -
~

400 --
-

Fig 3. Osteotomy gap bone

mineral content with time after
0 -- --T1-TT-r?-Tr-
the operation showing the ef-
fect of delayed application of
I 2 3 4 5 6 7 8 9 10 11 12 high rate micromotion (+ stan-
Weeks After Operation dard error of the mean,.n = 6).

DISCUSSION should be influenced by the rate of callus de-

formation. 0'Connor and Lanyonl7 showed
Mechanical influences are known to affect that in intact bone the change of strain rate is
the rate and pattern of fracture repair. Even an important variable influencing bone re-
short periods of applied mechanical stimula- modeling.
tion in experimental and clinical situations The optimal time when mechanical stimu-
have been shown to have significant effects lation is applied to the fracture either pas-
on bone healing.5,*39JlIn this applied me- sively or through weightbearing has not been
chanical stimulus, there exists the potential defined. Sarmiento et a121 suggested that
for fine tuning the mechanical signal to opti- early weightbearing induces rapid healing
mize the rate of healing. The fact that healing based on his studies in rats. It has been sug-
fractures have viscoelastic properties sug- gested that in fractures for which delay oc-
gests that strains produced in the callus curred before callus formation was encour-

Walking Stiffness Index

1
i

20

10

2 3 4 5
I
6
'
7
1
8
-
9
7
1
1
0
'
1
I

1
'
-,-,' 'I
,

I
--
,

1
I

' I
2
Fig 4. Walking stiffness index
with time after the operation
showing the influence of the ap-
plied micromotion rate (+ stan-
Weeks After Operation dard error of the mean, n = 6).
Number 3558
October, 1998 Mechanical Modulation of Fracture Healing S111

Walking Stiffness Index

-4-400 nmlraPnd (~lra

start)

2o I -I / I

Fig 5.time
with Walking
after stiffness
the operation
index lo ]-;7{7;7J;
showing the effect of delayed
0 T'-
application of high rate micro- -
motion (* standard error of the i! 3 4 5 6 7 8 9 1 0 1 1 1 2
mean, n = 6). Weeks After Operation

aged, there was a high chance of delayed
union (verbal communication, B. McKibben,
MD, 1989). In the current study, therefore,
the influences of strain rate and timing of
mechanical stimulation on healing were ex-
amined.
At the end of the 12-week allotted healing
period, the mechanical properties and the
rate of bone mineral content increase were
significantly higher in the 40 mm per second
group, but the rate of increase in the in vivo
walking fracture stiffness index was signifi-
cantly higher in the 400 mm per second
group. A low rate (2 mdsecond) of inter-
fragmentary displacement resulted in a sig-
nificantly reduced rate of osteotomy gap
mineralization and delay in restoration of
mechanical integrity. Thus the modulation of
indirect bone repair by very short periods of
daily cyclic mechanical stimulation is influ-
enced additionally by the rate at which the
interfragmentary movement is applied.
In terms of the postmortem mechanical re-
sponse of the isolated tibiae to torsional load-
ing, the 40 mm per second group had signifi-
cantly higher values of stiffness and strength
than the higher or lower displacement rate
groups. The postmortem tests used a pure tor-
sional loading pattern, whereas the in vivo
walking stiffness measurements subjected the
tibia to a more complex pattern of loading that
involved components of axial compression,
bending, and torsion. This more complex
physiological loading, together with the more

Torsional StHinesa Torsional Strength

(K oi control) (K o i control)

"O r
1-p c 0.01
x .. _______
T
*
~
~ ~~

i PNS.

I
100

1 .. a

Fig 6A-B. Mean (A) torsional

stiffness and (B) strength of
osteotomies 12 weeks after
11
surgery showing the influence zo
of the applied micromotion
rate (+ standard error of the , 0

s112 Goodship et al
Torsional Stiffness Torsional Strenglh
(% of control) (% of control)
70 T - T -__-__- 50
40
30
10
A ,,,,,,- lwrnm,scond
(ma ma*)
extensive periosteal bridging callus associ-
ated with the highest rate displacement group,
may explain the apparent discrepancy in me-
chanical data. This also would indicate that
the mode of loading used in assessment of
healing may influence the absolute values
used to determine an end point of healing.
Such a method of testing, which applies a pre-
dominantly nonphysiological loading, may
lead to an end point of healing that over or un-
derestimates the mechanical properties re-
quired for normal function.
In the second part of the study, applying
an osteogenic regimen of cyclic micromove-
ment in the latter half of the healing period
significantly reduced the rate of increase of
mineralization and fracture walking stiffness
index. It also decreased the torsional stiff-
ness and strength at the end of the 12-week
healing period compared with the group in
which the stimulus had been maintained
throughout the healing period.
The rationale for this study was to allow
periosteal bridging before imposing mechan-
ical stimulation to encourage adaptive hy-
pertrophy of the bony callus, thereby induc-
ing a consequent increase in mechanical
strength. However, the results showed that a
change imposed on the fracture’s mechanical
environment during the bridging phase of the
healing process induced a radical disruption
to the pattern and progression of healing.
This result may be explained by the fact that
Clinical Orthopaedics
and Related Research
T---T
p < 005
Fig 7A-B. Mean (A) tor-
sional stiffness and (B)
strength of osteotomies 12
weeks after surgery showing
the effect of delayed applica-
__ tion of high rate micromotion
Iwmnlsmnd looma;- (+ standard error of the
(latsrpft)
mean, n = 6).
the repair process, sensitive to the magnitude
and distribution of strain, responds to a
change by remodeling the callus to accom-
modate the strain characteristics of the new
mechanical environment. It also provides in-
direct evidence that the early phase of the re-
pair process is important in terms of sensitiv-
ity to biophysical stimulation.
In this early phase of the repair process, a
biofeedback mechanism appears to influence
the mechanical environment of the fracture
according to the degree of stabilization pro-
vided by the fixation device.14 It has been
shown that fixation with a rigid device re-
sults in high initial ground reaction forces
but a reduced rate of repair, whereas less
rigid stabilization leads to a lower initial
ground reaction force but an increased repair
rate and increasing ground reaction forces as
healing progresses.10
These data again may indicate the impor-
tance of an appropriate biophysical environ-
ment in the early stages of the repair process
to set the scene for normal healing. Inappro-
priate fixation may counter essential signals
required to initiate the repair process, lead-
ing to delayed union or nonunion. Indeed, it
is more likely that fracture repair is inhibited
by intervention, and that the optimal envi-
ronment for healing is achieved by removal
of inhibitory influences.
The indirect repair process consists of
two components, endochondral ossification
Number 3558
October, 1998
and intramembranous ossification, both of
which are influenced by interfragmentary
motion. Endochondral ossification proceeds
from the hematoma formed in the fracture
gap as a result of the initial damage. This
hematoma becomes organized, and a
process of differentiation begins in which
the connective tissue type changes as a func-
tion of time and environment. In this cas-
cade of connective tissue differentiation, the
entire spectrum of connective tissues is seen
from blood to bone through granulation tis-
sue, fibrous tissue, fibrocartilage, hyaline
cartilage, woven bone, and, ultimately,
lamellar bone. This bone formation results
from endochondral ossification of the hya-
line cartilage in the soft callus as a conse-
quence of vascular invasion. The process is
very similar to that of original bone forma-
tion in embryogenesis by endochondral os-
sification of the cartilage anlage.
Indirect bone repair also is associated
with a second component of osseous repair:
periosteal and endosteal bridging callus. In
this component, bone formation occurs by
intramembranous ossification. The amount
and distribution of the periosteal callus often
is greater than that of the endosteal callus,
possibly because of the physical constraints
of the medullary cavity, although an addi-
tional factor could be related to the mechani-
cal environment. The neutral axis of the bone
lies in the medullary cavity. Thus an exten-
sive formation of callus in this location
would not provide as great a contribution to
the restoration of mechanical integrity as the
periosteal callus. An extensive proliferation
of periosteal callus places loadbearing tissue
at a distance from the neutral axis where it
effectively can make a substantial contribu-
tion to resist bending and torsional loads. In
the current study, applying a short period of
cyclic movement early at a high rate was
shown to have a marked influence on the dis-
tribution of periosteal callus.
Activating the periosteum to form the
bridging callus involves initial proliferation
of the fibrous component of the periosteum,
Mechanical Modulation of Fracture Healing s113
which forms a fibrous bridging callus. This
is accompanied by proliferation of cells in
the osteogenic layer, which leads to in-
tramembranous ossification of the collage-
nous scaffold by direct mineralization. The
activation of this periosteal bridging callus is
maximal at the fracture line and diminishes
as a function of distance proximal and distal
to the fracture line. The mechanism for this
graded periosteal activation is not clear. The
bridging between fragments by periosteal ac-
tivation is initially by the fibrous layer of the
periosteum and can be imaged by ultrason-
agraphy in the first few weeks of the repair
process.20 Intramembranous ossification oc-
curs progressively from the area overlying
the bone fragment to the bridging callus until
bony union occurs.
After union, the areas of new bone form-
ing the total callus are remodeled to restore a
near normal organ and tissue architecture.
Data from this study and numerous previous
investigations indicate that repair process is
sensitive to biophysical and biologic manip-
~ l a t i o n . ' - ~ ~ ~This
* ~ - *sensitivity,
~ however,
has yet to be exploited fully to improve and
control the clinical management of fractures,
or perhaps more importantly to provide opti-
mal conditions for the progression of normal
repair. The transduction mechanisms by
which these biophysical stimuli influence
the processes of modeling and remodeling in
the intact and repairing bone also are not un-
derstood sufficiently to permit the use of ap-
propriate pharmacologic interventions in-
stead of controlled mechanical stimulation.
A previous study showed that a given
amount of strain applied at a higher rate will
produce a greater osteogenic response in in-
tact bone." It now is accepted that brisk
walking and a high rate of activity may help
to reduce bone loss in osteoporosis.23 Thus
experimental and clinical evidence suggest
that rate of deformation is a potent os-
teogenic signal. The current study suggests
that bone formation in fracture healing also
may respond to high deformation rates, so
the rate of fracture loading in active or pas-

S114 Goodship et al
sive stimulation may be indicated to enhance
fracture repair. It is possible that these high
rates of deformation may influence numer-
ous possible transduction pathways.
The endochondral component of the indi-
rect repair process involves numerous con-
nective tissues before the formation of bone
that are viscoelastic in terms of their material
properties. Thus when this component of the
callus is deformed at high rates, its stiffness
will be increased, consequently inducing
higher levels of deformation in the adjacent
cortical fragments. High strain rates in the
cortical bone of the fragments could activate
periosteal osteoblasts via osteocytic signal-
ing,22 thus inducing periosteal callus forma-
tion. In addition, the high rates of deforma-
tion may induce greater streaming potentials
associated with fluid flow in the fracture cal-
lus. This also may contribute to a greater cal-
lus formation, influencing the progression of
tissue differentiation in the soft callus. These
mechanical signals may contribute to activa-
tion of genes associated with vascularization
of the callus. In experimental studies of dis-
traction osteogenesis, in which regenerate
bone formation occurs by the process of in-
tramembranous ossification, an association
between mechanical distraction and subse-
quent activation of vascular endothelial
growth factor expression has been demon-
strated (verbal communication, T. A. Ein-
horn, MD, 1995).
The action of the cyclic mechanical stim-
ulus on the early callus therefore may also be
associated with the control of connective tis-
sue differentiation at the molecular level.
The sensitivity of the repair process, not only
to short periods of superimposed cyclic me-
chanical stimulation, but also to specific
characteristics of such a stimulus, which can
either enhance or inhibit healing, may have
significant clinical implications. Until the in-
teractions between biophysical stimuli and
cellular processes have been elucidated,
there is clear potential to monitor and control
bone repair by modulation of the mechanical
environment.
Clinical Orthopaedics
and Related Research
CLINICAL RELEVANCE
Earlier experimental and clinical studies have
shown that imposing small amounts of axial
micromovement for short periods each day
will affect healing. The inhibited healing of-
ten seen in experimental osteotomies or hu-
man fractures immobilized by external fixa-
tion frames can be prevented. The current
study shows that healing under these same
conditions is enhanced by the early applica-
tion of strain at a moderately high rate similar
to that obtained during brisk walking. Such
rapid loading, applied early after osteotomy,
helps rather than hinders healing. The studies
also show that imposing an axial micromove-
ment, similar to that used previously (0.5 mm
across a 3-mm gap), to an osteotomy that has
healed for 6 weeks will inhibit healing as
compared with early application of the same
stimulus from the start of healing. The late
application of an imposed mechanical stimu-
lus will reduce the effectiveness of this stim-
ulus in promoting healing.
Acknowledgments
The authors thank P. Walker for technical assis-
tance and Sarah Whitehouse for help with the sta-
tistical analysis.
References
1 . Aro HT,Kelly PJ, Lewallen DG, Chao EYS: The ef-
fects of physiologic dynamic compression on bone
healing under external fixation. Clin Orthop
256:260-273, 1990.
2. Claes L, Augat P, Suger G , Wilke HJ: Influence of
size and stability of the osteotomy gap on the suc-
cess of fracture healing. J Orthop Res 15577-584,
1997.
3. Claes LE, Wilke H-J, Augat P, et al: Effect of dynami-
sation on gap healing of diaphyseal fractures under
external fixation. Clin Biomech 10:227-374, 1995.
4. Cunningham JL, Evans M, Harris JD, Kenwright J:
The measurement of stiffness of fractures treated
with external fixation. Eng Med 16:229-232, 1987.
5. De Bastiani G, Aldegheri R, Renzi Brivio L: The
treatment of fractures with a dynamic axial fixator. J
Bone Joint Surg 66B:538-545, 1984.
6. Egger EL, Gottsauner-Wolf F, Palmer J, et al: Ef-
fects of axial dynamization on bone healing. J
Trauma 34:185-192, 1993.
7. Evans M, Kenwright J, Cunningham JL: Design and
performance of a fracture monitoring transducer. J
Biomed Eng 10:64-69,1988.
Number 3558
October, 1998
8. Goodship AE, Kelly DJ, Rigby HS, et al: The Ef-
fect of Different Regimes of Axial Micromove-
ment on the Healing of Experimental Tibia1
Fractures. In Bergmann G , Kolbel R, Rohlmann A
(eds). Biomechanics: Basic and Applied Re-
search. Dortrecht, Martinus Nijhoff 441446,
1987.
9. Goodship AE, Kenwright J: The influence of in-
duced micromovement upon the healing of experi-
mental tibial fractures. J Bone Joint Surg
67B:650-655,1985.
10. Goodship AE, Watkins PE, Rigby HS, Kenwright J:
The role of fixator frame stiffness in the control of
fracture healing: An experimental study. J Biomech
26:1027-1035,1993,
11. Kenwright J, Goodship AE: Controlled mechanical
stimulation in the treatment of tibial fractures. Clin
Orthop 241:3&48,1989.
12. Kenwright J, Richardson JB, Cunningham JL, et al:
Axial movement and tibial fractures: A controlled
randomised trial of treatment. J Bone Joint Surg
73B:654-659,1991.
13. Lanyon LE, Goodship AE, Pye CJ, MacFie JH: Me-
chanically adaptive bone remodelling. J Biomech
15:141-154, 1982.
14. Lazo-Zbikowski J, Aguilar F, Mozo F, et al: Bio-
compression external fixation: Sliding external os-
teosynthesis. Clin Orthop 206: 169-184, 1986.
15. MacGinitie LA, Wu DD, Cochran GV: Streaming
potential in healing, remodelling and intact cortical
bone. J Bone Miner Res 8:1323-1335.1993.
Mechanical Modulation of Fracture Healing s115
16. Noordeen MH, Lavy CB, Shergill NS, et al: Cyclical
micromovement and tibial fractures. J Bone Joint
Surg 77B:645-648, 1995.
17. O’Connor JA, Lanyon LE: The influence of strain
rate on adaptive bone remodelling. J Biomech
151767-781, 1982.
18. Perren SM: Physical and biological aspects of frac-
ture healing with special reference to internal fixa-
tion. Clin Orthop 136:175-196, 1979.
19. Perren SM, Cordey J, Rahn BA, et al: Early tempo-
rary porosis of bone induced by internal fixation im-
plants: A reaction to necrosis not to stress protection.
Clin Orthop 232139-151,1988.
20. Ricciardi L, Perissinotto A, Dabala M: Mechanical
monitoring of fracture healing using ultrasound
imaging. Clin Orthop 293:71-76, 1993.
21. Sarmiento A, Schaeffer JF, Beckermann L, et al:
Fracture healing in rat femora as affected by func-
tional weight bearing. J Bone Joint Surg
59A:369-375,1977.
22. Skeny TM, Bitensky L, Chayen J, Lanyon LE:
Early strain-related changes in enzyme activity in
osteocytes following bone loading in vivo. J Bone
Miner Res 4:783-788, 1989.
23. Smith EL, Gilligan C, McAdam M, et al: Deterring
bone loss by exercise intervention in premenopausal
and postmenopausal women. Calcif Tissue Int
441312-32 1, 1989.
24. Yamagishi M, Yoshimura Y The biomechanics of
fracture healing. J Bone Joint Surg 37A: 1035-1068,
1955.