Seumal of Ory
e602 Te eur of se Jl Sugey, 1
G'9%5 Ortop essa Sone)
Role of Muscles in Lumbar Spine Stability
in Maximum Extension Efforts
Mack Gardner-Morse, Ian A. F. Stokes, and “Jeffrey P. Laible
Depaviment of Orthopaedics ond Rehabilitation, and "Department of Civ and Environmental Hngincering,
University of Vermons, Burlington, Vermont, US.A.
‘Summary: Many problems of the lumbar spine that cause pain ave attributed to instability. The ligamentous
spine (without museles) is unstable at very low compressive loads. This study examined the hypothesis that
instability of the lumbar spine is prevented under normal cizeumstances by the stiffness of spinal musculature,
Without active responses from the neuromuscular control systems. The effect of muscle activity (Corce and
stiffness) on the stability of the lumbar spine was analyzed for maximum voluntary extension efforts with di
ferent spinal postures in the sagitial plane. The analysis included realistic three-dimensional representation
‘of the muscular snatomy with muscles erossing soveral motion scemonts. ‘The stiffness of motion segments
‘was represented using in vitro measured properties, Under a range of conditions with maximum extension
‘effort, active muscle stiffness was required to prevent the lumbar spine from buckling. The dimensionless
valve of the muscle stiffness parameter q as a function of activation and length had to be greater than a
critical value in the range of 37-4.7 in order to stabilize the spine. Experimentally determined values of ¢
ranged from 0.5 to 42. These analyses demonstrate how changes in motion segment stiffness, muscle activa
tion stratogy, or muscle stiffness (due to degenerative changes, injuries, fatigue, and so on) might lead to spinal
instability and “self-injury”
In vito, the ligamentous lumbar spine is unstable at
compressive loads of 88 N (8); however, in vivo the
compressive force acting on the lumbar spine exceeds
2,600 N, as demonstrated by measurements of intea-
discal pressure (25) and biomechanical models of lift
ing (32). Thus, the spinal column is inherently unstable,
but it is believed that in vivo a combination of muscle
forces and muscle stifnesses stabilize it. Spinal insta-
bility or segmental instability is a concept used clini-
cally to describe a situation in which motion segments
‘apparently are hypermobile (10). Although this clini-
cal condition is notoriously poorly defined (1), i gen.
erally is considered a degenerative condition of the
motion segments. Most biomechanical studies have
assumed that it is simple overload of the spine that
causes injuries (5) and their painful. consequences
Therefore, prevention of injury and effective rehabil-
itation has been directed primarily at minimizing spi-
nal forces. This viewpoint does not take into account
that muscles are required to stabilize the ligamentous
spine or the possible destabilizing effects of poor new
romuscular coordination,
Received Apri 2, 1994; acsoptod Decomber 22, 1994
‘Address correspondence and reprint requess fo M, Gatdnes=
Morse at Dopattmant of Orthopedics and Rehabilitation, Stat
ford Hall, University of Vermont, Burlington, VT 05405-0084,
USA,
Muscle and ligamentous forces increase the com-
pressive load on the spine, increasing its tendency to
buckle; however, the muscle forces are required to
stabilizé the spine. The relative magnitude of these
competing effects must be resolved with quantitative
analyses,
If muscles acted as pure force generators, then the
nervous system would be responsible for stabilizing
the spine by producing an appropriate response to
‘every small disturbance. Instead, it appeats that acti-
vated muscles act not only as force generators but also
as stabilizing springs (2.6.7). The amount of muscle
sfiffiness depends on the degree of muscle activation,
soit can be set or modulated in advance under control
‘of the nervous system. This regulation of muscle stiff-
ness by activation (and mote commonly coactivation.
of antagonist muscles) is used in targeted movements
to control the trajectory of a body segment (15).
Previous investigations of the stability of the spine
started with the observation that the ligamentous
spine is highly unstable (8.2133). Bergmark (2) exam-
ined the possibility that muscle stiffness stabilizes the
Jumbar spine for extension efforts using a model with
simplified muscle anatomy and motion segments The
distribution of muscle force was reduced to static de-
terminacy by a number of simplifying assumptions;
however, it was found that the muscle stiffness re-
quired for stability was high compared with values
802ROLE OF MUSCLES IN LUMBAR SPINE STABILITY 803
FIG. 1, Anteroposterior view of the lumber spine model. The
Wtergray cylinders represent the lumbar vertebrae, wi rigid
connections supporting muss alechmont, Me thera is «gid
‘ody to which all thoracic muscles aach. The postion of SL is
Indiated by a simple cylinder The rst of the sacrum and the
pelvis which were fixed) ace not shown. Active muscles (activity
fniresponcing to the maximum extension effort) are thown ae
epiadess with diameters proportional to ther physiologic ctoss-
feetional areas, and inactive muscles ae sown a ein ines.
subsequently reported by Crisco and Panjabi (6,7
from 2 review of the literature on muséle physiology.
This leaves open the question of how muscle stiffness
stabilizes the spine.
Crisco and Panjabi (7) explored the eritical mascle
stiffness required to maintain stability in five differ
cnt hypothetical muscle architectures attached to @
straight spine but did not establish how this would
apply to realistic anatomy or how the required sti
ness might relate o properties of active muscle.
Dietsich et al. (2) modeled the musculature of the
spine as solid finite elements with linear passive and
active stifness. The model buckled, but they did not
establish how such buckling is avoided in vivo.
Because of the approximations and simplifying as-
sumptions in previous studies, it has not been estab
Fished quantitatively whether muscle stiffness alone is
adequate to stabilize the spine in response to small
perturbations. We analyzed stability using realistic
data for spinal geometry, motion segment stiffness,
and musele anatomy. The overall hypothesis was that
damage occurs in the low back when muscles are not
sufficiently well coordinated to maintain both equilib-
rium and stability at all spinal levels We addressed
‘maximum extension efforts because they are consid-
ered more challenging to stability and have unique
solutions in the analysis of distributions of muscle
fotce. This avoided a major simplifying assumption of
previous studies We hypothesized that the lumbar
spine normally is stabilized in response to small dis-
turbances from the maximum effort state by the stiff
ness of spinal musculature. In this way, it would be
“self-stabilized” without active responses from the
neuromuscular control system, which has inherent
time delays.
METHODS
"The analysis of amber spins stability involved two sequential
steps: (a) an analysis of eistibation of msele Tore, whieh i
volved ince programming with equilibrium and physiotogc con
‘ents, wa wed to ealulate the nanimum extension effort and
corresponding motion segment and muscle foress, and (b) an i
fgenvalue bucking analysis of spinal stabilty was done aes, using
the motion segment and muscle foees found in the firs step
The geometry of the lumbar spine model was specified in teams
of vertebral body positions locations of muscle atachment points,
and phyeiologe cross sectional areas of muscle (3,34) (Hg. 1). The
dorsal musculature, rectus abdominis and psoas were represented
by 66symmerie muscle pars The ame thee simplifying seu
ns were made about al of the muscles (a) they take a sesght
Fine path from origin to inscrtion, (b) their contractile stress
(oreefmusee area) was betwoon 0 (uselosprodoce only tension)
and 469 kPa (334), and (¢) they have active stifiness proportional
te activation and inversely proportional to muscle length (2). The
model used beam clement (11) matehed to the experimentally
derived data on linear slfiness given by Panjabi etal. 27) to
represent the spinal motion segments The thorsx and sacsumipel-
vs were oth considered as rg bodies ‘The sacrumipelvs was