estrogen depletion is associated with decreases in compact bone

advertisement
ESTROGEN DEPLETION IS ASSOCIATED WITH DECREASES IN COMPACT BONE VISCOELASTIC PROPERTIES
+*Les, C; *Vance, J; *Christopherson, G; *Patel, B; **Turner, A; *Fyhrie, D
+*Henry Ford Hospital, Detroit, MI. 313.916.3166, Fax: 313.916.8064, les@bjc.hfh.edu
13
0.06
12
0.04
Sham
OVX
11
δ
tanδ
E1, GPa
Introduction: Compact bone is a viscoelastic material (2,10). Using
constant strain-rate experiments, Carter and Hayes (1), determined that the
Young’s modulus of trabecular bone is proportional to strain rate to a small
(0.06) power. Whether this relationship changes significantly with age,
disease, nutrition, or training in compact bone has not, to our knowlege, been
addressed.
Bone loss associated with estrogen depletion is associated with increased
fracture risk (3). While much of the bone loss seen in postmenopausal
osteoporosis is from the cancellous envelope, significant losses are also seen
in compact bone (4). Standard screening methods (e.g., DXA, QCT, QUS)
may not be effective at identifying population at risk of fracture (5). If
changes in viscoelastic behavior of the bone material also accompany estrogen
depletion, then fracture risk may be increased dramatically without necessarily
decreasing the bone mineral density that is often measured by these imaging
modalities. Moreover, we have shown (6) that short-term (1y) estrogen
depletion can result in subtle but structurally-significant changes in the
distribution of bone density within the compact bone of the diaphysis,
possibly altering the direction of bending under normal as well as abnormal
loads. A combination of material redistribution and changes in viscoelasticity
could well result in significant structural changes without concomitant overall
bone loss.
One means of evaluating the viscoelastic properties of a material uses
subyield oscillatory tests at a range of frequencies (2,7). If an oscillatory
stress σ is applied to a linear viscoelastic material at an angular frequency ω:
ε=ε0 cos (ω
ωt-δ
δ)], where
[σ=σ
σ=σ0 cos ωt], the resulting strain ε can be defined as [ε=ε
δ is the phase angle between σ and ε. The complex modulus of the material
under these conditions, E*, can be defined as [E*=E1+iE2], where E1 is the
real or storage modulus (equivalent to the Young’s modulus): [E1=(σ
σ0/εε0) cos
σ0/εε0) sin δ]. The
δ], and E2 is the imaginary, dynamic, or loss modulus: [E2=(σ
loss tangent, or tanδ, is defined as [tanδ
δ=E2/E1], and is a measure of how
effectively the material can damp an oscillatory stress.
The objective of this study was to evaluate the viscoelastic properties of
compact bone material as a function of anatomic site and long-term estrogen
depletion. We hypothesized that the dependence of E1 and tanδ on stress
frequency, and the anatomic variation in these material properties, would be
altered with long-term estrogen depletion.
Methods: Under IACUC approval, 5yo Warhill sheep were ovariectomized
(OVX, N=6) or subjected to a sham surgery (N=6). Three years later, ewes
were sacrificed and the left radius/ulna (the two bones fuse early in life)
harvested and stored at –20C. Six 2x2x19mm (craniocaudal x lateromedial x
proximodistal) beams were cut from each radial diaphysis (craniomedial,
cranial, craniolateral, caudomedial, caudal, and caudolateral sectors) under
cold water irrigation, and stored in 0.9% saline solution at –20C. The beams
were thawed and tested in 0.9% saline solution at 37C in a Perkin Elmer
DMA7e dynamic mechanical analyzer. Each beam was tested in 3-point
bending (outer supports 15mm apart), in craniocaudal orientation. A static
load of 550mN and dynamic load of 500mN was applied in a frequency scan
from 1 to 20Hz at 0.2Hz intervals. E1 and tanδ were calculated at each
frequency. The plot of E1 as a function of frequency for each test (averaged in
triplicate) was fit to an exponential model (E1 = a*freqb). The coefficient a,
exponent b, and tanδf (f=1,3,6,9,12,15,18,20Hz) were used as dependent
variables in a repeated-measures ANOVA, using treatment (OVX or sham)
and anatomic location as categorical variables.
Results: Significant changes in both a and b were associated with long-term
estrogen depletion (a: Sham=10.466∀.0133SE, OVX=10.521∀.0136, p=.016;
b:Sham=.0439∀.00429, OVX=.0237∀.00438, p=.009; Fig.1). The two
parameters were constant across anatomic sites (p=.415). At frequencies
>3Hz, OVX was associated with a decreased tanδ (p<.004; Fig.2). At
frequencies<12Hz, there was a tendency (.047>p>.001) for the caudal and
cranial sectors to have higher values for tanδ than the other sectors (Fig.3).
This distribution did not appear to change with OVX (.615>p>.054).
Sham
0.02
10
OVX
0.00
0
10
0
20
10
20
Frequency, Hz
Frequency, Hz
Fig.1 (Left) Storage modulus as a function of frequency, ∀1SE. OVX was
associated with a decrease in the sensitivity of E1 to oscillation frequency.
Fig.2 (Right) Tanδ as a function of frequency, ∀1SE. OVX was associated, at
frequencies>3Hz, with a decrease in the damping function.
1-9Hz
12-20Hz
Cranial
Lateral
0.06
0.03
0.00
Cranial
Tanδ
δ
Lateral
1Hz
3Hz
6Hz
9Hz
0.06
0.03
0.00
Tanδ
δ
12Hz
15Hz
18Hz
20Hz
Fig.3: Tanδ (radius) as a function of anatomic position (angle) and test
frequency. At low frequencies, there was more effective damping in the
cranial and caudal sectors. No change in this distribution with estrogen
depletion was demonstrated.
Discussion: The stiffness of compact bone in this system becomes
dramatically less sensitive to changes in stress rate (>45% decrease in the
stress-rate exponent b) after three years of estrogen depletion. The damping
characteristics of the material, particularly at higher frequencies, deteriorated
as well. Each of these changes alone could result in substantial alterations in
the dynamic properties of the structure. Together, these losses in viscoelastic
properties could easily help explain an increase in fracture risk without
concomitant local or global changes in bone mineral density.
We were able to demonstrate a significant anatomic variation in the lowfrequency damping characteristics of the bone material, with the more
effectively-damping material found in the regions subjected to substantial
compression and tension during the sagittal bending experienced by this bone
in normal locomotion (8), and supporting an hypothesis of material
distribution in compact bone to enhance bending in certain preferred
directions(9). However, no such variation was found in the stress-rate
sensitivity, and did we demonstrate change in the anatomic variation in
damping with OVX, as we did with mineral density (6). This would suggest
that these viscoelastic properties, and possible changes in them, may be
determined by more global mechanisms that are not necessarily accounted for
by classical osteonal remodeling.
References: 1)Carter, JBJS 59A:954, 1977. 2)Garner, JBiomechEngr
122:166,2000.3)Wasnich, in Primer on Metabolic Bone Diseases
(Favus,ed):249, 1996. 4) Bell, JBMR 14:111,1999. 5)Melton, AnnIntMed
112:516, 1990. 6) Les, ProcORS 26:477, 2001. 7)Findley: Creep and
relaxation of nonlinear viscoelastic materials:90, 1976. 8)Lanyon, JBiomech
12:93,1979. 9)Les, JBiomech 30:355,1997. 10) McElhaney, JApplPhys
21:1231,1966.
**Colorado State University, Ft.Collins, CO.
48th Annual Meeting of the Orthopaedic Research Society
Paper No: 0089
Download