Full text of DOI:10.1359/jbmr.2002.17.9.1613

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 17, Number 9, 2002
© 2002 American Society for Bone and Mineral Research
Low-Magnitude Mechanical Loading Becomes Osteogenic
When Rest Is Inserted Between Each Load Cycle
SUNDAR SRINIVASAN,1 DAVID A. WEIMER,2 STEVEN C. AGANS,1 STEVEN D. BAIN,3
and TED S. GROSS1
ABSTRACT
Strategies to counteract bone loss with exercise have had fairly limited success, particularly those regimens
subjecting the skeleton to mild activity such as walking. In contrast, here we show that it is possible to induce
substantial bone formation with low-magnitude loading. In two distinct in vivo models of bone adaptation, we
found that insertion of a 10-s rest interval between each load cycle transformed a locomotion-like loading
regime that minimally influenced osteoblast activity into a potent anabolic stimulus. In the avian ulna model,
the minimal mean (؉SE) periosteal labeled surface (Ps.LS) observed in the intact contralateral bones (1.6 ؎
1.5%) was doubled after 3 consecutive days of low-magnitude loading (3.8 ؎ 1.5%; p ؍ 0.03). However,
modifying the regimen by inserting 10 s of rest between each load cycle significantly enhanced the periosteal
response (21.9 ؉ 4.5%; p ؍ 0.03). In the murine tibia model, 5 consecutive days of 100 low-magnitude loading
cycles did not significantly alter mean periosteal bone formation rate (BFR) compared with contralateral
bones (0.011 ؎ 0.005 ␮m3/␮m2 per day vs. 0.021 ؎ 0.013 ␮m3/␮m2 per day). In contrast, separating each of
10 of the same loading cycles with 10 s of rest significantly elevated periosteal BFR (0.167 ؎ 0.049 ␮m3/␮m2
per day; p ؍ 0.01). Endocortical bone formation parameters were not altered by any loading regimen in either
model. We conclude that 10 s of rest between each load cycle of a low-magnitude loading protocol greatly
enhances the osteogenic potential of the regimen. (J Bone Miner Res 2002;17:1613–1620)
Key words: mechanotransduction, exercise, osteoblast, bone mass, osteogenic
to a fate of elevated fracture risk for a substantial portion
their lives.(6)
INTRODUCTION
IMINISHED BONE mass in the elderly arises because of
degradation of the normal balance between bone re-
Mechanical loading of bone has substantial potential to
induce bone formation.(7,8) However, application of exercise
regimens in humans as a means of augmenting bone mass has
met with mixed results.(9) The elderly, in particular, are unable
to comply consistently with the high-impact strenuous loading
events typically associated with even minimal bone accre-
tion.(10) Unfortunately, lower-magnitude exercise strategies are
extremely limited in their ability to enhance bone mass.(11)
A growing body of experimental data support the hypoth-
esis that fluid flow near osteocytes underlies mechanotrans-
duction within bone. Osteocytes reside within lacunae and
canaliculi inside the mineralized matrix.(12–14) Mechanical
D
sorption and bone formation observed in the young
adult.(1,2) Current treatment strategies have focused on in-
hibiting bone resorption and are highly successful, particu-
larly in blocking elevated resorption associated with meno-
pause.(3) However, options for elevating bone formation as
a means of enhancing bone mass are extremely limited.(4,5)
Given our rapidly aging population, the current inability to
augment safely and effectively bone mass consigns millions
The authors have no conflict of interest.
1Orthopedic Science Laboratories, Department of Orthopedics and Sports Medicine, University of Washington, Seattle, Washington,
USA.
2Department of Orthopedics, University of Cincinnati, Cincinnati, Ohio, USA.
3SkeleTech, Inc., Bothell, Washington, USA.
1613

1614
SRINIVASAN ET AL.
loading greatly accentuates fluid flow and diffusion within
bone(15,16) and osteocytes are highly responsive to fluid
flow.(17,18) Additionally, bone formation induced by me-
chanical loading is spatially associated with locations of
large strain gradients, which, in turn, stimulate fluid
flow.(19,20) However, given the spatial dimensions of the
osteocyte’s lacunae and its cellular glycocalix coating, flow
of fluid past the cell body and cell processes is likely to be
highly viscous.(21–23) This physical consideration suggests
that if bone is loaded repetitively (i.e., at typical stride
frequencies of 1 Hz and higher), sufficient time may not
exist for osteocyte level fluid flow to recover from inertial
damping effects between each load cycle.(24) Fluid flow and
thus osteocyte stimulation, therefore, potentially are much
reduced after the first few cycles of repetitive loading. If so,
it is likely that inserting a “sufficient” time period between
each loading cycle would heighten fluid flow near osteo-
cytes by facilitating recovery from inertial damping effects.
Therefore, we hypothesized that inserting a nonloaded
rest pause between each loading cycle would increase the
osteogenic potential of a low-magnitude locomotion-like
mechanical loading regimen. We examined our hypothesis
using both the avian ulna model and a recently developed
noninvasive murine tibia model of bone adaptation. Specif-
ically, we contrasted the ability of low-magnitude loading
regimens with and without a 10-s rest interval inserted
between each load cycle to induce osteoblastic activation
and/or bone formation.
MATERIALS AND METHODS
Avian ulna model
In this study, eight middle-aged male turkeys (2 years
old) underwent the functionally isolated avian ulna proce-
dure.(25) The left ulna of each turkey was isolated from
functional loading by parallel metaphyseal osteotomies per-
formed on both ends of the bone. Transcutaneous Stein-
mann pins were passed through stainless steel caps placed
over the exposed diaphyseal bone ends to enable external
loading via a pneumatic actuator. After surgery, the turkeys
were assigned randomly to one of two groups (standard
regime or rest-inserted regime). Both groups underwent
1-week protocols. Beginning on day 3 postoperatively, the
ulna was loaded externally in bending, with a strain distri-
bution similar to that induced during moderate wing flap.
For the standard regime (n ϭ 4), the left ulna was loaded for
100 cycles with a 1-Hz sawtooth waveform on days 3, 4,
and 5. Animals undergoing the rest-inserted regime (n ϭ 4)
underwent an identical waveform (same magnitude and
loading rate), except that a 10-s unloaded rest pause was
inserted between each of the 100 loading cycles (Fig. 1).
Mineralization of osteoid associated with bone apposition at
the ulnar middiaphysis was detected via administration of
calcein (3 mg/kg intravenously [iv]) on day 6.
FIG. 1. The (A) standard waveform and (B) rest-inserted waveform
were identical in magnitude and total cycle number in the avian ulna
study (illustrated via middiaphyseal normal strains). They differed only
by the 10-s pause inserted between each loading cycle of the rest-
inserted waveform. The first 12 s of each regimen is illustrated. (C) For
the murine tibia study, single load cycles for the low- and high-
magnitude groups are illustrated via applied load waveforms. The
rest-inserted load waveform was identical to that of the low-magnitude
group, with 10 s of rest inserted between each cycle.
scope. Using mean end loading conditions established via
two separate strain-gauged calibration bones,(26) beam the-
ory was applied to the middiaphysis cross-section geometry
of each intact contralateral bone to quantify peak middi-
aphyseal strains induced in both experimental groups. Custom
written software (PV-Wave; VNI, Inc., Boulder, CO, USA)
was used to detect the extent of periosteal calcein labeling.
Percent periosteal-labeled surface (Ps.LS) was determined as
the ratio of Ps.LS to the middiaphyseal periosteal perimeter.
After death, 200-␮m-thick sections were removed from
the middiaphysis of both the experimental and the intact
contralateral ulnas. The sections were hand ground to 80
␮m, blinded, and imaged (100ϫ) using a spot digital cam-
Noninvasive mouse tibia model
Eighteen adolescent female C57Bl/6J mice (10 weeks)
era mounted on a Nikon diaphot epifluorescence micro- underwent external loading of the right tibia using the

REST BETWEEN LOAD CYCLES
1615
bending in the medial-lateral plane by applying force to the
lateral distal tibia metaphysis via a computer-controlled
linear force actuator. Mice were assigned randomly to three
groups: (1) low magnitude (low, n ϭ 6), (2) low magnitude
rest-inserted (rest-inserted, n ϭ 6), and (3) high magnitude
(high, n ϭ 6). The low-magnitude group underwent 100
cycles/day of a 0.25 N peak load, 1 Hz, 0.01/s strain rate
trapezoidal waveform for 5 consecutive days (inducing 650
␮⑀ peak normal strain at the middiaphysis in separate cal-
ibration mice). For the low-magnitude rest-inserted group,
peak load (0.25 N) and peak strain rate (0.01/s) were equiv-
alent to the low-magnitude group. However, instead of 100
daily load cycles, the rest-inserted mice were exposed to a
10-cycle/day regime with a 10-s pause inserted between
each load cycle. The high-magnitude group underwent a
100-cycle/day 1-Hz regime with the peak load and strain
rate doubled compared with the low-magnitude group (0.5
N; 0.02/s) and served as a positive control for an osteogenic
response in the model (Fig. 1). The overall daily external
loading period for the mice in all groups was 100 s. All
animals were allowed 23 days of free cage activity after the
5-day loading regimen to facilitate consolidation of new
bone. All mice received calcein injections (15 mg/kg intra-
peritoneally [ip]) on day 1 and day 26 to detect osteoblast
activity over the course of the experiment. After death, on
day 28, 100-␮m-thick middiaphyseal cross-sections were
taken from the right (experimental) and left (intact con-
tralateral) tibias and were mounted unstained for evaluation
of cross-sectional areal properties and dynamic indices of
bone formation. Sections were imaged (100ϫ) with identity
blinded. On both the endocortical and periosteal surfaces,
single-labeled surface (sLS), double-labeled surface (dLS),
and interlabel thickness (Ir.L.Th) were measured. From
these data, mineralizing surface (MS; MS ϭ (dLS ϩ sLS/
2)/bone surface [BS] ء 100), mineral apposition rate (MAR;
MAR ϭ Ir.L.Th/interlabel time period [Ir.L.t]), and surface
referent bone formation rate (BFR; BFR ϭ MAR ء MS/BS)
were calculated.(27) Static measures of the periosteal enve-
lope (periosteal area [Ps.Ar]) and endocortical envelope
(endocortical area [Ec.Ar]) were used to determine cross-
sectional cortical area (Ct.Ar). Animal-specific peak middi-
aphyseal strains induced at the initiation of the loading
experiments were calculated using beam theory to apply end
loading conditions established in separate calibration exper-
iments via a combined strain gauging and finite element
(FE) analysis approach.
FIG. 2. (A) Composite fluorescent micrograph of the ulna middi-
aphysis illustrating the minimal Ps.LS induced by a 1-week application
of the standard waveform. (B) In contrast, substantial Ps.LS was
evident on ulna loaded using the rest-inserted protocol (arrows high-
lighting calcein labeling). (C) Both the low-magnitude standard wave-
form (p ϭ 0.03) and the rest-inserted waveform (p ϭ 0.03) significantly
elevated Ps.LS versus intact contralateral bones (*). Ps.LS induced by
rest-inserted loading was, in turn, significantly greater than that induced
by the standard waveform (p ϭ 0.03, ∧).
Statistics
Nonparametric statistical comparisons were used to as-
sess within-group and across-group treatment effects. Wil-
coxon paired tests (p ϭ 0.05, one-tailed) were used to
distinguish differences between loaded bones and intact
contralateral bones. Kruskal-Wallis tests (p ϭ 0.05, two-
tailed) were used to identify differences due to different
loading regimes, with Mann-Whitney tests used as a post
noninvasive murine tibia loading device.(26) With the mouse
anesthetized (2% isoflurane), the device secured the proxi-
mal tibia metaphysis from motion via a brass gripping cup.
The tibia diaphysis then was placed under “cantilever” hoc follow-up (p ϭ 0.05, two-tailed).

1616
SRINIVASAN ET AL.
TABLE 1. MEAN (ϮSE) ENDOCORTICAL BONE FORMATION DATA FROM NONINVASIVE LOADING OF THE MOUSE TIBIA
Ec.MS (%)
Ec.MAR (␮m/day)
Ec.BFR (␮m3/␮m2 per day)
Low
High
Rest
Intact
Load
Intact
Load
Intact
Load
51.5 (Ϯ6.18)
61.0 (Ϯ8.72)
45.3 (Ϯ3.23)
37.3 (Ϯ5.50)
45.0 (Ϯ5.06)
40.2 (Ϯ5.72)
0.85 (Ϯ0.17)
0.82 (Ϯ0.11)
0.71 (Ϯ0.08)
0.85 (Ϯ0.12)
0.80 (Ϯ0.12)
0.87 (Ϯ0.13)
0.45 (Ϯ0.10)
0.55 (Ϯ0.13)
0.33 (Ϯ0.05)
0.33 (Ϯ0.06)
0.36 (Ϯ0.06)
0.36 (Ϯ0.07)
No significant differences were noted.
Low, low-magnitude loading group; high, high-magnitude loading group; rest, rest-inserted loading group; intact, intact left tibia; load,
experimentally loaded right tibia.
high-magnitude and rest-inserted loading were more stim-
ulatory for osteoblasts than low-magnitude loading, no sta-
tistical differences were observed between high-magnitude
and rest-inserted loading for any bone formation parameter.
As compared with intact bones, both high-magnitude (p ϭ
0.05) and rest-inserted (p ϭ 0.03) loading induced small but
significant increases in Ct.Ar (5.1% and 6.2%, respec-
tively). Ct.Ar was achieved primarily via expansion of the
periosteal envelope (Table 2).
RESULTS
Avian ulna model
The loading regimens induced nearly identical mean (Ϯ
SE) peak normal strains in the standard waveform group
(820 Ϯ 50 ␮⑀) and the rest-inserted group (830 Ϯ 60 ␮⑀;
p ϭ 0.40). Only two of eight intact contralateral ulnas
showed any periosteal labeling (both in the rest-inserted
group). Averaged across all turkeys, 1.6 Ϯ 1.5% of the
periosteal surface was labeled in the intact bones. The
standard loading waveform elicited a small but significant
elevation in the Ps.LS compared with intact contralateral
bones (3.8 Ϯ 1.5%; p ϭ 0.03; Fig. 2). The rest-inserted
waveform was highly osteogenic (21.9 Ϯ 4.5%; p ϭ 0.03).
The 5.8-fold increase in labeled surface elicited by the
rest-inserted loading compared with standard loading was
statistically significant (p ϭ 0.03).
DISCUSSION
We used two distinct in vivo models of bone adaptation to
examine whether insertion of a brief rest interval between
each load cycle enhances the osteogenic potential of low-
magnitude mechanical loading. In the avian ulna model, the
periosteum was quiescent under normal conditions. We
found that low-magnitude locomotion-like loading only
minimally influenced osteoblast activity. Surprisingly, in-
Noninvasive mouse tibia model
Peak longitudinal normal strains were equivalent in the sertion of a 10-s rest interval between each load cycle
low-magnitude and rest-inserted groups (680 Ϯ 20 ␮⑀ vs. transformed the low-magnitude regimen into a potent ana-
665 Ϯ 25 ␮⑀; p ϭ 0.39) and were doubled in the high- bolic stimulus. Using the noninvasive murine tibia-loading
magnitude group (1330 Ϯ 50 ␮⑀). The young mice exam- model, we found that low magnitude did not alter periosteal
ined in this study exhibited active osteoblast activity on both bone formation versus normal control bones. When a 10-s
endocortical and periosteal surfaces at the initiation of the rest interval was inserted between each loading cycle, peri-
study. None of the three loading regimens altered endocor- osteal bone formation was significantly enhanced, despite a
tical bone formation parameters compared with normal con- 10-fold reduction in daily load cycle number. Last, we
tralateral tibia (Table 1). In contrast, osteoblastic activity on found that the bone formation elicited by 10 cycles of
the periosteum was significantly enhanced by both the high- rest-inserted loading was statistically comparable with that
magnitude and the rest-inserted regimes (Fig. 3). Histolog- induced by a standard locomotion-like waveform 10-fold
ically, new bone formation was lamellar, with no evidence greater in cycle number and 2-fold greater in load magni-
of woven bone. Compared with paired intact tibia, perios- tude and rate.
teal MS was not significantly altered by low-magnitude
The ability of rest-inserted loading to significantly aug-
loading (Fig. 4) but was significantly enhanced by high- ment the osteogenic potential of low-magnitude loading in
magnitude (p ϭ 0.04) and rest-inserted loading (p ϭ 0.01). both adult and growing skeletons suggests promise as a
Likewise, periosteal MAR was not altered by low- potential strategy to build bone mass via exercise. However,
magnitude loading but was significantly elevated by both a number of facets of this observation remain to be resolved
high-magnitude (p ϭ 0.01) and rest-inserted loading (p ϭ before its successful application. First, the described studies
0.01; Fig. 4). As a result, periosteal BFR was significantly used models of cortical bone adaptation. Augmentation of
elevated in both high-magnitude (p ϭ 0.01; 3.7-fold vs. both trabecular and cortical bone would enhance most ef-
paired contralateral tibia, 14.5-fold vs. low magnitude) and fectively bone structural properties, but it is not known
rest-inserted load groups (p ϭ 0.01; 2.3-fold vs. paired whether rest-inserted loading will be similarly osteogenic
contralateral tibia, 8.0-fold vs. low magnitude) but was not for trabecular bone. To date, we have not attempted to
altered in the low-magnitude group (Fig. 4). Although both optimize the rest interval. Our initial choice of a 10-s rest

REST BETWEEN LOAD CYCLES
1617
FIG. 4. Periosteal bone formation parameters at the tibia midshaft in
response to low-magnitude, high-magnitude, and low-magnitude rest-
inserted loading. Low magnitude loading did not significantly alter (A)
MS, (B) MAR, or (C) BFR compared with the intact contralateral tibia.
Both high-magnitude and rest-inserted loading significantly enhanced
these parameters (*). In addition, both high-magnitude and rest-inserted
loading induced significantly greater mineral apposition and bone for-
mation compared with the low-magnitude loading (∧). No statistical
differences were observed between high-magnitude and rest-inserted
loading.
pause appears somewhat serendipitous because based on
our preliminary reports,(28,29) a recent study using larger-
magnitude loading protocols (about fourfold greater peak
strain magnitude than the low-magnitude protocols used
here) found that rest intervals of 3.5 s and 7 s were inef-
fective, but 14 s increased bone formation compared with
repetitive loading regimes.(30) It remains to be determined
whether optimal rest intervals are similar across species,
different peak strain magnitudes, and different cycle dura-
tions. Also, it remains to be assessed whether rest-inserted
loading will ultimately augment structural properties suffi-
ciently to decrease fracture risk, because the initial studies
described here assessed the ability of rest-inserted loading
to initiate an osteogenic response (i.e., periosteal activation
FIG. 3. Composite fluorescent micrographs of the mouse midshaft
illustrate (A) minimal periosteal bone formation in response to the
low-magnitude regimen, and (B) substantial periosteal bone formation
stimulated by the high-magnitude and (C) rest-inserted regimens (ar-
rows). Endocortical bone formation was similar for each group and no
different from that observed in the intact contralateral tibia.

1618
SRINIVASAN ET AL.
TABLE 2. MEAN (ϮSE) CROSS-SECTIONAL AREAL DATA FROM NONINVASIVE LOADING OF THE MOUSE TIBIA
Ps.Ar (mm2)
Ec.Ar (mm2)
Ct.Ar (mm2)
Low
High
Rest
Intact
Load
Intact
Load
Intact
Load
0.889 (Ϯ0.036)
0.880 (Ϯ0.018)
0.840 (Ϯ0.022)
0.870 (؎0.023)
0.857 (Ϯ0.026)
0.898 (Ϯ0.022)
0.399 (Ϯ0.025)
0.385 (Ϯ0.017)
0.347 (Ϯ0.013)
0.352 (Ϯ0.013)
0.390 (Ϯ0.014)
0.403 (Ϯ0.015)
0.491 (Ϯ0.013)
0.495 (Ϯ0.006)
0.493 (Ϯ0.012)
0.518 (؎0.011)
0.466 (Ϯ0.012)
0.495 (؎0.009)
Significant differences are in bold.
Low, low-magnitude loading group; high, high-magnitude loading group; rest, rest-inserted loading group; intact, intact left tibia; load,
experimentally loaded right tibia; bold, loaded tibias significantly increased versus intact tibias ( p Ͻ 0.05).
and bone formation). Extension of this response over weeks tude(7,8) and rate,(37,38) loading frequency,(39) and fluid
and months likely will be required to significantly enhance flow(40,41) have been shown potently osteogenic. In general,
structural properties.
these investigations have established the paradigm that the
Although the data of this study do not specifically address greater the mechanical stimulus, the greater the response
the mechanism by which rest-inserted loading derived its elicited from bone cells or bone tissue. Comparison of the
potency, they circumstantially support the thesis that fluid standard low- and high-magnitude groups confirms this
flow near osteocytes mediates mechanotransduction within pattern of response in the murine tibia model. However, in
bone. In this context, we recently reported a numerical the growing mouse, we found that 10 cycles of low-
model of lacunocanalicular fluid flow, suggesting that load- magnitude rest-inserted loading substantially enhanced
induced fluid flows near osteocytes were reduced substan- bone formation compared with control bones, while 100
tially (Ͼ45%) beyond the first few load cycles of repetitive cycles of an identical-magnitude repetitive loading without
locomotion-like mechanical loading.(24) Our physical inter- rest between cycles did not alter bone formation versus
pretation of this phenomenon was related to the forced flow control bones. In addition, the amount of bone formation
of viscous interstitial fluids through the fiber-filled canalic- induced by 10 cycles of daily rest-inserted loading was not
ular annuli and the consequent “inertia” in the fluid flow. In significantly different from that achieved by the positive
the present in vivo studies, tissue deformations were iden- control high-magnitude group exposed to twice the load
tical for both standard locomotion-like loading and rest- magnitude and rate and 10 times the number of load cycles.
inserted loading; yet, insertion of a rest interval activated a The ability of rest-inserted loading to induce more bone
substantially larger proportion of the periosteum (turkey) formation (vs. low-magnitude loading) or equivalent bone
and induced more bone formation (mice). These results formation (vs. high-magnitude loading) with less actual
suggest that rest-inserted loading may derive its efficacy loading of bone (cycle number and/or magnitude, respec-
from enhancing secondary stimuli induced when bone is tively) directly challenges the concept that high-magnitude,
loaded mechanically (i.e., stimuli other than deformation of high-impact loading is required to build bone mass.
the tissue). It is reasonable to assume that a 10-s rest interval
In this context, the broad relevance of these data lie with
would diminish inertial fluid flow effects associated with their potential to extend substantially the skeletal benefits of
locomotion-like exercise that contains nonloaded periods mild exercise. The ability to induce bone formation in
only during the swing phase of gait. Thus, fluid flow near growing or mature skeletons via low-magnitude nonimpact
osteocytes (and associated processes such as shear stress, loading would make bone-building exercise accessible to all
nutrient exchange, and convection of cytokines) likely were ages while minimizing injury risk. A number of possibilities
elevated in the rest-inserted regimes.
might be explored readily to implement rest-inserted load-
Alternatively (or additionally), the effectiveness of rest- ing. Tai Chi, which already has been suggested to enhance
inserted loading may arise from other distinct processes. For balance in the elderly,(42) might be modified such that pos-
example, a substantial body of literature indicates that re- tures between movements are deliberately held for longer
petitive stimuli rapidly saturate cellular responses in various periods of time. Similarly, hand weights might be held in
tissues(31–33) and that the response of bone to a few cycles of fixed positions between each load repetition to enhance
repetitive loading is similar to that achieved with orders of bone mass in the upper extremities. With such ease of
magnitude more load cycles.(25,34,35) Recent in vivo data implementation, rest-inserted waveforms also might be
confirm that bone becomes accommodated to repetitive combined with the low-magnitude high-frequency loading
loading, as separation of a 360-cycle daily loading protocol that is highly osteogenic for trabecular bone(39) as a means
into 4 bouts of 90 cycles over the course of a day enhances of focally enhanced bone structure at any skeletal site.
the anabolic response of bone to mechanical loading.(36)
In summary, insertion of a 10-s rest interval between each
Although rest-inserted loading remains to be optimized, it load cycle transformed a low-magnitude minimally osteo-
represents a departure from established modalities of repet- genic loading regimen into a potent osteogenic stimulus in
itively loading bone tissue. In both in vivo and in vitro two distinct in vivo models of bone adaptation. We found
models, mechanical stimuli such as increased strain magni- that 10 cycles of low-magnitude loading per day inter-

REST BETWEEN LOAD CYCLES
1619
11. Pruitt LA, Taaffe DR, Marcus R 1995 Effects of a one-year
high-intensity versus low-intensity resistance training program
on bone mineral density in older women. J Bone Miner Res
10:1788–1795.
spersed with 10 s of rest induced equivalent bone formation
as a regimen of 100 cycles/day of twice the strain magnitude
and twice the strain rate. These studies with cortical bone
emphasize that it should be possible to augment bone struc-
ture without subjecting an individual to high-magnitude
high-impact loading. If so, low-magnitude or mild exercise
regimes interspersed with rest intervals between load cycles
may prove a safe and highly effective treatment for both
acute (e.g., space induced) and chronic (e.g., paralysis or
aging induced) bone loss.
12. Doty SB 1981 Morphological evidence of gap junctions be-
tween bone cells. Calcif Tissue Int 33:509–512.
13. Palumbo C, Palazzini S, Marotti G 1990 Morphological study
of intercellular junctions during osteocyte differentiation.
Bone 11:401–406.
14. Knothe Tate ML, Steck R, Forwood MR, Niederer P 2000 In
vivo demonstration of load-induced fluid flow in the rat tibia
and its potential implications for processes associated with
functional adaptation. J Exp Biol 203(Pt 18):2737–2745.
15. Piekarski K, Munro M 1977 Transport mechanism operating
between blood supply and osteocytes in long bones. Nature
269:80–82.
ACKNOWLEDGMENTS
16. Zeng Y, Cowin SC, Weinbaum S 1994 A fiber matrix model
for fluid flow and streaming potentials in the canaliculi of an
osteon. Ann Biomed Eng 22:280–292.
17. Klein-Nulend J, Semeins CM, Ajubi NE, Nijweide PJ, Burger
EH 1995 Pulsating fluid flow increases nitric oxide (NO)
synthesis by osteocytes but not periosteal fibroblasts—
correlation with prostaglandin upregulation. Biochem Biophys
Res Commun 217:640–648.
The authors thank Chris Jerome, Ph.D., for helpful dis-
cussions regarding statistical analyses. This study was
funded by The Whitaker Foundation for Bioengineering and
the National Institute of Arthritis, Musculoskeletal and Skin
Diseases grant AR48102 (to T.S.G.) and the University of
Cincinnati Orthopaedic Research and Education Foundation
(to S.S. and D.A.W.).
18. Cheng B, Zhao S, Luo J, Sprague E, Bonewald LF, Jiang JX
2001 Expression of functional gap junctions and regulation by
fluid flow in osteocyte-like MLO-Y4 cells. J Bone Miner Res
16:249–259.
19. Gross TS, Srinivasan S 1998 Association of osteoblastic acti-
vation with spatial strain gradients. Tran Orthop Res Soc
44:944.
20. Qin YX, McLeod KJ, Guilak F, Chiang FP, Rubin CT 1996
Correlation of bony ingrowth to the distribution of stress and
strain parameters surrounding a porous-coated implant. J Or-
thop Res 14:862–870.
21. Cane V, Marotti G, Volpi G, Zaffe D, Palazzini S, Remaggi F,
Muglia MA 1982 Size and density of osteocyte lacunae in
different regions of long bones. Calcif Tissue Int 34:558–563.
22. Michel CC 1988 Capillary permeability and how it may
change. J Physiol 404:1–29.
REFERENCES
1. Bass S, Delmas PD, Pearce G, Hendrich E, Tabensky A,
Seeman E 1999 The differing tempo of growth in bone size,
mass, and density in girls is region-specific. J Clin Invest
104:795–804.
2. Hansen MA, Overgaard K, Riis BJ, Christiansen C 1991 Role
of peak bone mass and bone loss in postmenopausal osteopo-
rosis: 12 year study. BMJ 303:961–964.
3. Hosking D, Chilvers CED, Christiansen C, Ravn P, Wasnich
R, Ross P, McClung M, Balske A, Thompson D, Daley M,
Yates AJ 1998 Prevention of bone loss with alendronate in
postmenopausal women under 60 years of age. N Engl J Med
338:485–492.
4. Sogaard CH, Mosekilde L, Thomsen JS, Richards A, McOsker
JE 1997 A comparison of the effects of two anabolic agents
(fluoride and PTH) on ash density and bone strength assessed
in an osteopenic rat model. Bone 20:439–449.
5. Finkelstein JS, Klibanski A, Arnold AL, Toth TL, Hornstein
MD, Neer RM 1998 Prevention of estrogen deficiency-related
bone loss with human parathyroid hormone-(1–34): A ran-
domized controlled trial. JAMA 280:1067–1073.
6. Cummings SR, Black DM, Nevitt MC, Browner WS, Cauley
JA, Genant HK, Mascioli SR, Scott JC, Seeley DG, Sherwin
R, Steiger P, Vogt TM 1990 Appendicular bone density and
age predict hip fracture in women. The Study of Osteoporotic
Fractures Research Group [see comments]. JAMA 263:665–
668.
7. Rubin CT, Lanyon LE 1985 Regulation of bone mass by
mechanical strain magnitude. Calcif Tissue Int 37 (4):411–7.
8. Turner CH, Forwood MR, Rho JY, Yoshikawa T 1994 Me-
chanical loading thresholds for lamellar and woven bone for-
mation. J Bone Miner Res 9:87–97.
9. Prince RL, Smith M, Dick IM, Price RI, Webb PG, Henderson
NK, Harris MM 1991 Prevention of postmenopausal osteopo-
rosis. A comparative study of exercise, calcium supplementa-
tion, and hormone-replacement therapy. N Engl J Med 325:
1189–1195.
23. Zhu W, Lai WM, Mow VC 1991 The density and strength of
proteoglycan-proteoglycan interaction sites in concentrated so-
lutions. J Biomech 24:1007–1018.
24. Srinivasan S, Gross TS 2000 Canalicular fluid flow induced by
bending of a long bone. Med Eng Phys 22:127–133.
25. Rubin CT, Lanyon LE 1984 Regulation of bone formation by
applied dynamic loads. J Bone Joint Surg Am 66:397–402.
26. Gross TS, Srinivasan S, Liu C, Clemens TL, Bain SD 2002
Non-invasive loading of the murine tibia: An in vivo model for
study of mechanotransduction. J Bone Miner Res 17:493–501.
27. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H,
Meunier PJ, Ott SM, Recker RR 1987 Bone histomorphom-
etry: Standardization of nomenclature, symbols, and units.
Report of the ASBMR Histomorphometry Nomenclature
Committee. J Bone Miner Res 2:595–610.
28. Srinivasan S, Gross TS 2000 Intermittent rest enhances osteo-
blastic activation induced by mechanical loading. Trans Or-
thop Res Soc 25:628.
29. Srinivasan S, Weimer DA, Liu CC, Bain SD, Gross TS 2001
The osteogenic potential of rest-inserted loading. Trans Orthop
Res Soc 26:325.
30. Robling AG, Burr DB, Turner CH 2001 Recovery periods
restore mechanosensitivity to dynamically loaded bone. J Exp
Biol 204(Pt 19):3389–3399.
31. Davies PF, Dewey CF Jr, Bussolari SR, Gordon EJ, Gimbrone
MA Jr 1984 Influence of hemodynamic forces on vascular
endothelial function. In vitro studies of shear stress and pino-
cytosis in bovine aortic cells. J Clin Invest 73:1121–1129.
10. Nelson ME, Fiatarone MA, Morganti CM, Trice I, Greenberg
RA, Evans WJ 1994 Effects of high-intensity strength training
on multiple risk factors for osteoporotic fractures. A random-
ized controlled trial. JAMA 272:1909–1914.

1620
SRINIVASAN ET AL.
32. Chitwood RA, Jaffe DB 1998 Calcium-dependent spike- 40. Reich KM, Gay CV, Frangos JA 1990 Fluid shear stress as a
frequency accommodation in hippocampal CA3 nonpyramidal
mediator of osteoblast cyclic adenosine monophosphate pro-
neurons. J Neurophysiol 80:983–988.
duction. J Cell Physiol 143:100–104.
33. Cordoba-Rodriguez R, Moore KA, Kao JP, Weinreich D 1999 41. Klein-Nulend J, van der Plas A, Semeins CM, Ajubi NE,
Calcium regulation of a slow post-spike hyperpolarization in
vagal afferent neurons. Proc Natl Acad Sci USA 96:7650–
7657.
Frangos JA, Nijweide PJ, Burger EH 1995 Sensitivity of
osteocytes to biomechanical stress in vitro. FASEB J 9:441–
445.
34. Unemura Y, Ishiko T, Yamauchi T, Kurono M, Mashiko S 42. Wolfson L, Whipple R, Derby C, Judge J, King M, Amerman
1997 Five jumps per day increase bone mass and breaking
force in rats. J Bone Miner Res 12:1480–1485.
35. Turner CH 1998 Three rules for bone adaptation to mechanical
stimuli. Bone 23:399–407.
P, Schmidt J, Smyers D 1996 Balance and strength training in
older adults: Intervention gains and Tai Chi maintenance [see
comments]. J Am Geriatr Soc 44:498–506.
36. Robling AG, Burr DB, Turner CH 2000 Partitioning a daily
mechanical stimulus into discrete loading bouts improves the
osteogenic response to loading. J Bone Miner Res 15:1596–
1602.
37. O’Connor JA, Lanyon LE, MacFie H 1982 The influence of
strain rate on adaptive bone remodelling. J Biomech 15:767–
781.
38. Mosley JR, Lanyon LE 1998 Strain rate as a controlling
influence on adaptive modeling in response to dynamic load-
ing of the ulna in growing male rats. Bone 23:313–318.
39. Rubin C, Turner AS, Bain S, Mallinckrodt C, McLeod K 2001
Address reprint requests to:
Ted S. Gross, Ph.D.
Department of Orthopedics and Sports Medicine
University of Washington
325 Ninth Avenue, Box 359798
Seattle, WA 98104-2499, USA
Anabolism. Low mechanical signals strengthen long bones. Received in original form October 8, 2001; in revised form Feb-
Nature 412:603–604. ruary 26, 2002; accepted March 12, 2002.