APPLIED SCIENCES

Biodynamics

Contributions of lower 'extremity joints to

energy dissipation during landings
SONG-NING ZHANG, BARRY T. BATES. and JANET S. DUFEK
Exercise Science Unit, Biornechanics/Sports Medicine Laboratory, The Universily of Tennessee, Knoxville, Knoxville, TN
37996; and Universiry of Oregon, Eugene, OR

ABSTRACT
ZHANG, S-N.. B. T. BATES, and I. S. DUFEK. Contributions of lower extremity joints to energy dissipation during landings. Med.
Sci. Sports Exerc., Vol. 32, No. 4, pp. 812-5 19,2000. Purpose: The purpose of the study was to investigate changes in lower extremity
joint energy absorption for different landing heights and landing techniques. Methods: Nine healthy, active male subjects volunteered
to perform step-off landings from three different heights (0.32 m, 2.5 m-'; 0.62 m, 3.5 m^: and 1.03 m, 4.5 m") using three different
landing techniques (soft. SFL; normal. NML, and stiff landing, STL). Each subject initially performed five NML trials at 0.62 m to
serve as a baseline condition and subsequently executed five trials in each of the nine test conditions (3 heights X 3 techniques).
Results: The results demonstrated general increases in peak ground reaction forces, peak joint moments, and powers with increases
in landing height and stiffness. The mean eccentric work was 0.52. 0.74, and 0.87 J-kg-' by the ankle muscles, and 0.94, 1.31, and
2.15 J-kg-' by the hip extensors, at 0.32,0.62, and 1.03 m, respectively. The nvernge eccentric work performed by the knee extensors
was 1.21, 1.63. and 2.26 ].kg-' for the same three heights. Conclusions: The knee joint extensors were consistent contributors to energy
dissipation. The ankle plantarflexors contributed more in the STL landings. whereas the hip extensors were greater contributors during
the SFL landings. Also a shift from ankle to hip strategy was observed as landing height increased. Key Words: STRATEGIES,
MECHANICAL DEMAND. MUSCULAR WORK

uring foot contact with ground in vigorous locomo-
tion, the body experiences tremendous impact
forces. Maximum vertical ground reaction force
values as high as 14.4 times body weight (BW) have been
reported (20) for single-leg landings from a double 'back
somersault. Stacoff et al. (25) showed that the first peak (Fl)
of the vertical component of ground reaction force (GRF)
ranged from 1000 to 2000 Newtons (N), whereas the second
peak (F2) values ranged from 1000 to 6500 N in landings
after a volleyball block jump. McNitt-Gray (17) demon-
strated that the maximum vertical ground reaction forces for
trained gymnasts were 3.9, 6.3. and 11.0 times BW for
landing from heights of 32. 72, and 128 cm, respectively.
Accumulation of high impact forces may pose a threat to the
integrity of the lower extremity and related overuse injuries
are often direct consequences of these impacts (22,23).
Many of these injuries are often associated with the knee-
joint structure and reported as common in running (16),
0 195-9131/00/3204-0sl a 0
MEDICINE & SCIENCE LN SPORTS & EXERCISEo
Copyright t3 2000 by the American College of Sports Medicine
Submitted for publication February 1998.
Accepted for publication September 1998.
8 12
basketball (19), tennis (24), football (15). volleyball (l2),
skiing (10). and the triathlon (3). Fifty-eight percent of all
injured female basketball players were engaged in landing
from a jump at their time of injury (13) and 40% of high-
level volleyball players experienced knee problems during
their playing careers (1 1). James et al. (16) reported knee
pain as the most common problem for runners.
The studies on biomechanical behaviors of the lower
extremity in landing have been focused on prediction of
impact forces (9). comparisons of landing techniques (4),
effects of landing velocities (18). and manipulations of
landing distances, heights, and techniques (8). Dufek and
Bates (8) demonstrated greater peak moments of the prox-
imal extensors (i.e., hip extensors) compared with the distal
extensors (i.e., plantarflexors). Few studies, however, inves-
tigated the contributions of various lower extremity muscle
groups to the total energy dissipation. DeVita and Skelly (4)
suggested that the ankle joint plantar flexors absorbed more
energy in a stiff landing, whereas the hip and knee extensors
absorbed more energy in the soft landing. McNitt-Gray (18)
demonstrated that elite gymnasts dissipated more energy
.r with ankle and hip extensors at the higher height compared
with their recreational counterparts, whereas the latter group
adjusted their strategy by increasing hip flexion and by
 C
, ,. .)nging the landing phase. Different views still exist in
literaturc regarding the various contributions made by
tllc. lower extrcrnity muscle groups to the reduction of im-
pact and energy. The femur and tibia are among the longest
bones in human body and significantly greater loading at the
b e e joint are expected. High mechanical output by the knee
musculature should be experienced consequently. Research
in this area, however. still fails to agree on the issue. There-
fore. the purpose of the study was to investigate changes in
?..?r_ryabsorption of lower extremity joints for different
I:. ding heights and landing techniques.
METHODS
+
Nine active men (ages: 25 5 yr, mass: 74.4 2 6.3 kg)
volunteered to participate as subjects in the study. Each
subject signed an Informed Consent Form approved by the
Institutional Review Board at the University of Oregon,
:,hich was consistent with the human subject policy of the
.merican College of Sports Medicine before the experi-
. mental sessions. All subjects were actively involved in
recreational sports on a regular basis (at least 2-3 times per
week) and had participated in sporting activities requiring
jumpingllanding skills. They were free of injury or other
physical impairment in the lower extremity at the time of
testing.
The subjects participated in two test sessions. The first
lest session was used to familiarize the subjects with the
:xperimental protocol and to obtain mean maximum knee
flexion angles for the nine test conditions that were combi-
nations of soft (SFL), normal (NML), and stiff (STL) land-
ing techniques (stiffness) and three landing heights (0.32 m,
0.62111,and 1.03 m). The subjects were asked to perform five
landing trials in each of the 10 conditions and the right knee
flexion angles were monitored using an electrogoniometer
(Penny and Gile,,1000 Hz). The mean maximum knee
flexion angles obtained were used to determine the range of
maximum knee flexion angle for each of the nine test
conditions for each subject with following equation:
%npc = ffmax 2 90
where a,,,, = range of maximum knee flexion angle and
a, = mean maximum knee flexion angle. The 9' criterion
was derived from the data of maximum knee flexion of
normal landings from a 60-cm height and two standard
deviations were applied to ensure proper range of variation
(5). The a,,, values thus obtained for each subject at each
height X technique combination (condition) were used to
monitor landing technique during the second test session.
Any trial with a maximum knee flexion angle beyond % ,,
was discarded and an additional landing was performed.
The second session began with an anthropometric mea-
surement portion and was followed by the activity testing
protocol. All subjects were encouraged to actively warm up
before the testing. Subjects performed five step-off landing
trials from a raised platform in each of the 10 conditions for
a total of 50 trials. The landings were initiated with right
foot leading forward, the left foot remaining in contact with
LANDING BIOMECHANICS
the platform, and thc ccnter ot' grnvity as stcady as possible
before the step-off. The subjects were instructed to keep
both hands on their buttock so that their upper limbs were
parallel to the sagittal plane and clear from obstructing their
hip reflective markers during landing. Among the 10 con-
ditions, the first was a base-line condition performed at the
beginning with normal landing from the median height (i.e.,
0.63 m). Subsequently, the subjects were asked to perform
landings in nine test conditions which were a combination of
SFL, NML, and STL landing trials from the three landing
heights of 0.32,0.63, and 1.03 m representing three contact
velocities (i.e., 2.5,3.5, and 4.5 mas-'). The order of heights
was from the lowest to the highest to minimize potential risk
of injury. The landing techniques were randomized in such
a way that the soft and stiff landing occurred randomly
either before or after normal landing at each height.
The right sagittal view was recorded using two high-
speed video cameras (200 Hz, Motion Analysis Corpora-
tion, Santa Rosa. CA) with one camera focusing on the'
performance of lower extremity and the other camera on the
total body. The lower extremity view was mapped on to the
total body view to obtain more accurate measurement of
lower extremity kinematics. Retro-reflective markers were
placed on the right side of the ear canal, acromion, lateral
humeral epicondyle, radial styloid process, greater trochani
ter, lateral femoral epicondyle, lateral malleolus, heel, and
head of the fifth metatarsal. Two force platforms (AMTI)
were used to collect data of ground reaction forces (GRF)
and moments of force for left and right limbs. Only the right
side signals were used for further analyses. The electrogo'-
niometer was placed on the right knee joint to monitor the
maximum knee flexion angles and landing stiffness (tech-
nique-defined as maximum knee flexion angle) during the
test sessions. The force platform and electrogoniometer
signals were sampled at 1000 Hz using a biomechanical
system (Ariel Dynamics Inc., Trabuco Canyon, CA). The
synchronization between the video, force platform, and
electrogoniometer systems was accomplished through a
light-emitting-diode which was present in the video views,
triggered by the force platform and fed as an analog channel
to the Ariel system. The video images were digitized to
obtain coordinates of the joints using the commercial soft-
ware (Motion Analysis). The data of the kinematic coordi-
nates, ground reaction forces, and moments were imported
into customized software to compute typical joint kinemat-
ics, segmental inertia properties (14), and joint kinetics via
an inverse dynamics model (26).
The variables evaluated included the range of motion
(ROM) of the three lower extremity joints, peak GRF,peak
joint moments and powers, and total eccentric work per-
formed by the different lower extremity muscle groups.
Critical values were assessed with statistical procedures, and
the time-history of the kinematic and kinetic variables was
examined qualitatively. The ground reaction forces and joint
kinetic values were normalized to the body mass for each
individual subject while the representative moment and
power curves were normalized to landing phase; the latter
was defined as the time from foot contact to the minimal
Medicine & Science In Sports & Exercise@ 8 13
.ABLE 1. Collapsed group means and SD of maximum GRF and selected ROM across techniques or heights.
F1 F9
Condition (N.kgI1) (N.ktl)
32 8.27' 25.14'
62
(2.58)
17.30'
(8.52)
32.17b
(4.55) (8.27)
103 30.60e 46.50'
(6.07) (10.24)
SFL 16.48d 29.17"
NML
(10.48) (1 1.99)
STL 21 .Oaf 41.45'
(10.34) 11 1.661
Values in parenthesis are standard deviation (SD); Significant difference level: p < 0.05; see text for full description ot the variables.
a Significantly different between 32 and 62 cm.
' Significantly different between 62 and 103 cm.
CSignificantly different between 103 and 32 cm.
dSignificantly different between SFL and NML.
*Significantly different between NML and STL.
'Significantly different between STL and SFL.
vertical center of gravity (COG) position during landing. By
convention, both the positive and negative values for joint
moment would represent extensor and flexor moment, and
the positive and negative power values would indicate en-
ergy generation and absorption. A three-factor 3 X 3 X3
(height X technique X joint) univariate analysis of variance
(ANOVA) was computed for the total work performed by
lower extremity muscle goups. Two-factor 3 X 3 ANOVAs
(height X technique) were performed on the peak GRF,
ROM, and moment and power variables. Because the peak
moments and powers do not occur at the same time and do
not fully reflect the total effort involved in the energy
reduction, only the height and technique were adopted as the
factors without consideration of the joint in the analyses.
Unless mentioned otherwise, the significant level for differ-
ence was set at P < 0.05.
RESULTS
Ground reaction force magnitudes. The typical
time-history curve of the vertical ground reaction force
demonstrated two distinctive maximums with first peak (Fl)
related to toe contact and second peak (F2)associated with
heel contact (6,7). A summary of the selected critical events
collapsed by height across techniques and by technique
across heights is provided in Table 1. The 3 X 3 ANOVA
(height X technique) results for F1 and F2 demonstrated
significant omnibus F (P < 0.0001). significant main effects
for height and technique (P < 0.0001), and insignificant
height X technique interactions (P > 0.05). Due to the lack
of interactions, the data were collapsed across three tech-
niques for each height and across three heights for each
technique (Table 1). Similar combinations were also made
and presented for the ROM, peak moment, and power vari-
ables. The peak GRF values (F1 and F2) increased signifi-
cantly (P C 0.05) from the lower to higher heights. When
comparing these values under.the increased landing stiffness
conditions, the significantly augmented peaks were ob-
served as well (Table 1). These results demonstrated a trend
8 14 Official Journal of the American College of Sports Medicine
Ankle ROM (.) Knee ROM (4) Hip ROM (.)
30.24' 52.02' 40.22"
(p)
3 .99
(21.02)
56.74"
(26.75)
49.86'
(6.82) (19.11) (26.94)
34.42' 63.61 ' 65.35'
(7.74) (1 7.42) (24.25)
34.38 77.Md 79.26"
(6.72)
28.69'
(6.751
of the increased loading to the body with increases in either
landing height or landing stiffness.
Joint ROM. A change of landing technique (stiffneps) is
practically defined as a change of lower extremity joint
angles during landings. More specifically stiffness was de-
fined as a change of maximal knee joint flexion angle during
the landing phase in this study. The 3 X 3 ANOVA
(height X technique) results indicated highly significant
omnibus F ( P < 0.0001), and significant main effects for
height and technique (P < 0.0001) for the hip, knee, and
ankle ROM. The hip and ankle ROM demonstrated insig-
nificant interactions between height and technique, whereas
the interaction for the knee ROM was significant (P <
0.01). The collapsed mean ROM values for the lower ex-
tremity joints generally increased with enhancements in
landing height except the changes in the ankle ROM from
62 to 103 cm (Table 1). Meanwhile the decreases in the
ROM with increases in landing stiffness (technique) were
seen across all three joints except the changes in the ankle
joint ROM from SFL to NML. The ankle joint ROM main-
tained small changes from SFL to STL and the most drops
in ROM at all three heights were accounted for mainly by
the hip and knee joint.
Kinetics. Mean representative curves of hip, knee, and
ankle joint moment and power normalized to landing phase
for NML at 0.62 m are provided in Figure 1 and 2. Two
distinctive peaks of first (AM 1 and Khl 1) and second (AM2
and KM2) maximum moment were present in both ankle
and knee joint moment curves (Fig. 1). Only one significant
peak was observed in hip joint moment (HM) during the
impact phase of landing. Similarly. two minimums were
observed for the power curves for the ankle (APl and AP2),
and knee (KP1 and KP2) joint, whereas only one minimum
was observed from the hip (HP) joint (Fig. 2). The maxi-
mum and minimum values in the moment and power curves
1 occurred approximately at TI and T2, indicating their close
associations with the first and second impact forces.
Peak extensor moments and powers represent maximum
efforts by certain muscle groups in energy absorption. The
 -2.0 .
0 20 40 60 80 100
(a) Percent
Figure 1-Representative average joint moment curves of (a) hip joint
moment, (b) knee joint moment, and (c) ankle joint moment, for
normal landing from 0.63 m. First and second vertical lines corre-
sponding to TI and T2 of GRF. See text for full description of the
labels.
3 X 3 ANOVAs (height X technique) were performed for
these variables. Significant effects were noticed for the
omnibus F (P < 0.0001), the main effects (P< 0.0001), and
insignificant interactions between height and technique (ex-
cept KM2),The values of AM1 and AP1 were significantly
lower than the values of AM2 and AP2 for all conditions,
indicating minimum effect of ankle plantarflexors in the
impact absorption at TI (Tables 2 and 3). However, the knee
extensors became increasingly involved in the absorption of
LANDING BIOMECHANICS
40 60
Percent
.,
Figure %Representative average joint power curve of (a) hip joint
power, (b) knee joint power, and (c) ankle joint power, for normal
landing from 0.63 m. First and second vertical lines corresponding to
T1 and T2 of GRF. See text for full description of the labels.
the first impact force at the two higher heights. The peak hip
moment and power that occurred later than AM1 and KM1
demonstrated significantly greater values than that of peak
ankle and knee extensor moments and power (Table 2 and
3). The timings for those peaks were consistent and arranged
in 'h sequential manner (Figs. 1 and 2).
The mean absolute and relative eccentric work is pro-
vided in Table 4. A 3 X 3 X 3 ANOVA (height X tech-
nique X joint) was conducted on the eccentric work
Medicine & Science in Sports & Exercise* 8 15
 ,
TABLE 2. Collapsed group means and SD of maximum joint moments across techniques or heights.
Condition AM1 AM2 KM1 KM2 . HM
32 0.57. 1.54' 1.73d 2.04' 4.1ga -6
(0.17) (0.38) (0.55) (0.38) (1.33) 6;
62 1.07O 2.06b , 3.91b 2.64b 5.8~~ 4...
103

SFL
(0.55) (0.55) (2.1 5) (0.64) (2.49)
NML 1.106 1.96' 4.10' 2.75' 6.216
(0.61) (0.61) (2.34) (0.74) (2.70)
STL 1.33' 2.40' 4.42' 2.94' 6.80'
(0.65) (0.69) (2.29) (0.84) (2.70)
Peak moment unit is in N.m.kg-,; values in parenthesis are standard deviation (SD); significant difference level: p < 0.05; see text for full description of the variables.
a Significantly different between 32 and 62 cm.
bSignificantly different between 62 and 103 cm.
CSignificantly different between 103 and 32 cm.
"Significantly different between SFL and NML.
'Significantly different between NML and STL.
'Significantly different between STL and SFL.

performed by the lower extremity. The result demonstrated
significant omnibus F, significant main effects for height,
technique and joint, and significant two-way interactionsfor
joint X height and joint X technique (P < 0.01). The
interaction between height and technique was not significant
(P > 0.05) and no significant three-way interactions were
found. Two 3 X 3 ANOVAs (joint X height and joint X
technique) were conducted as follow-up analyses. The in-
teractions between joint and height as well as joint and
technique are shown in Figure 3. The absolute mean values
sors showed decreased contributions (45.3% to 20.9%) to
the total energy dissipation as the landing stiffness in-
creased. The increases in landing heights did not influence
the relative contributions of each of the three lower extrem-
ity muscle groups consistently (Table 4). Clear anticipation
by the subjects was demonstrated by the joint positions at
contact; these data suggested that the initial ankle joint
positions remained constant while the initial knee and hip
joint angles demonstrated overall decreases as the landings
became stiffer. t

of eccentric work during the landing phase displayed the

trends of change that are slightly different from the relative
(percent) eccentric work (Table 4). For absolute work, the
ankle muscle group predominantly demonstrated increased
work with the increases in landing heights and slight im-
provement of eccentric work from SFL to STL. The hip and
knee joint extensor groups showed general decreased ec-
centric efforts with increased stiffness and improved work
values with increased landing heights (Table 4, Fig. 3). For
The purpose of the study was to examirie changes in
lower extremity joint energy absorption for different landing
heights and landing techniques. The changes in the landing
heights and the landing techniques are closely related to the
changes in mechanical demands placed on human body
' during the landing activities. The effects of mechanical

the relative eccentric work, however, the knee joint main- demands placed on the lower extremity and the correspond-
tained consistent and high mechanical outputs in all landing ing mechanical responses of lower extremity musculature
conditions. Meanwhile, the ankle muscles demonstrated in- during landing can be examined by evaluating changes in
creased contributions (12.5% to 38.1%) and the hip exten- landing techniques in relation to landing heights and the

TABLE 3. Collapsed group means and SD of minimum joint powers across techniques or heights.
Condition AP1 AP2 KP1 KP2 HP

103

SFL

NML

STL
(3.45) (3.47) (16.22) (7.85) (19.63)
s parenthesis are standard deviation (SD); significant difference level: P < 0.05; See text for full description of the variables.
Peak power unit is in W.kgml; ~ l u e in
'Significantly different between 32 and 62 cm.
bSignificantly different between 62 and 103 cm. I
cSi~nificantlydifferent between 103 and 32 cm.
dSlgnificantly different between SFL and NML.
a Significantly different between NML and STL.
'Significantly different between STL and SFL.

8 16 Official Journal of the American College of Sports Medicine

5~;
..;lean absolute and relal'~eeccentric work at individual lower extremity joints.
--
Work Percent Work
condition Ankle Knee Hlo Total Ankle Knee Hip
-7%~ 3.7SJ." 12.9 46.3 40.9
82 NML 2.78b,d 21.0 47.1 32.0
32 STL
62 SFL
62 NML
62 STL
103 SFL
103 NML
103 STL
/

Ecce !;lc work unit is in Jakg-': values in parenthesis are standard deviation (SD); significant difference level: p < 0.05; see text for full description of the variables.
rgifiereflt from NML at the same height.
bgifferent from STL at the same height.
~Qifferentfrom SFL at the same height.
dDiffereflt from 62 cm height with the same technique.
#Different from 103 cm height with the same technique.
'Different from 32 crn height with the same technique.

interaction of these two factors. The results suggest that several investigators that biarticular muscles are used for
th e are general increases in biomechanical responses with power transportation during locomotion (2,21). Prilutsky
th.; increased landing heights. However, the two peak and Zatsiorsky (21) demonstrated that the direction of
ground reaction forces responded differently with such
changes. The average and relative increases with respect to
the lowest height (0.32 m) were 136% for F1 and 44% for 7
F2 with all three landing techniques, indicating a more
Newtonirtn~mechanicalresponse from the subjects for F l
than for F2 without much neuromuscular intervention. The
average times from the contact to F1 and F2 were 9.6 and
2 L.2 ms. The F2 responses occurred much later after the first t

impact and therefore were considered more neuromuscular,

allowing more time for the lower extremity muscles to
absorb the impact. DeVita and Skelly (4) observed a pre-
contact muscular activity during landing, which suggests a
preset central neuromuscular program. This anticipation
may enhance the tension of the musculotendinous structures
in the foot and ankle, and other lower extremity regions
before the occurrence of landing contact so that greater energy SFL NML STL
lissipation may happen immediately after the impact. (a)
As mentioned previously, the mechanical demands in-,
creased with an increase in landing heights. All three lower
extremity extensor groups responded with an increase in
peak and total mechanical output for the same technique
(Tables 2-4). However, these changes were not undertaken
in similar fashion among the different lower extremity mus-
cle groups. Although the eccentric work performed by the
ankle plantarflexors was increased slightly, the increases in
the work done by the knee and hip extensors was more
apparent as the height changed from 0.32 to 1.03 m (Table
4). The absolute work data from this study suggested that
the ankle muscle group was less capable of energy absorp-
tion compared with hip and knee muscle groups, especially
at the two higher heights. The proximal muscles of the lower 32 62 103
extremity tend to be larger in volume than the distal mus-
(b)
Figure &Interactions of R0R.I between (a) technique and joint*and
cles; this is mainly due t~ greater cross-sectional areas (271,
longer muscle fibers, and relatively shorter tendons (I). This
*,
(h) height and joint. H,nnkle joint; knee joint; A. hip joint. For (a),
no significant differences were found among techniques for ankle joint;
mangement reduces the moment of inertia relative to the the differences between knee and hip using SFL was not signflcant;
q e differences between hip and ankle using STL was not signifiant;
hip joint and allows the leg to move more efficiently while all other comparisons were significant. For (b), significant differences
it also limits the capacity of energy absorption and genera- found between different combinations of joint and height (P <
tion in more distal muscle groups. It has been suggested by 0.05).

L A N D I N G BIOMECHANICS Medicine & Science in Sports & E x e r c i s e 8 17

power transfcr during landing was from distal to proximal
segments: it is therefore logical to assume that the energy
generated before landing contact can be transported from
the distal end to the more proximal and massive muscle
group for further dissipation during impact. An indirect
evidence from this study indicated that the peak powers
reached at different times with the second peak knee power
and the peak hip power occurring at an later time than the
second peak ankle power (Fig. 2).
The subjects displayed distinct strategies at the different
landing heights. In other words. each joint and muscle group
played significantly different roles of energy reduction in
the various mechanical loading situations (Fig. 2). It was
noted that the ankle maintained a stable mechanical output.
whereas the eccentric work by the knee and hip joint ex-
tensors was reduced with changes of technique from the soft
to stiffer landings at each height. As a result of these
changes, the total work was decreased significantly due to
the reduced time for energy reduction in normal and stiff
landing. The knee and hip extensors were equally important
in the landings with more lower extremity flexion (Table 4).
As the landing became stiffer, their ability in energy reduc-
tion decreased in different fashions. The hip extensors were
less effective than the knee muscles in the STL condition at
the low height. The same muscle group became more in-
volved in the energy reduction at 0.62 m and 1.03 m due to
the increases in their mechanical advantages in STL with
greater hip joint flexion.
The present study presented a different energy dissipation
pattern than has been previously reported in the literature.
The eccentric work by knee and ankle muscles was greater
for the soft landings compared with the stiff landings in a
study reported by DeVita and Skelly (4). In the current
experiment, significant increases in knee and hip work and
decreases in ankle work were observed in the soft landing
condition. McNitt-Gray (18) reported high hip and knee .
work. and small ankle work for landings from medium and
high heights (0.72 and 1.28 m) for recreational athletes
using a normal landing technique. Although we observed
similar changes in the knee and hip work, we saw an
increase in the ankle eccentric work as the landing height
increased. It is logical to a g u e that with the increased lower
extremity flexion and longer landing phase at higher
heights. the ankle musculature should perform more work.
One possible explanation for the discrepancies may lay in
the different activities involved in the present and previous
studies. DeVita and Skelly investigated soft and stiff Iand-
ings from a 0.59 m height, whereas McNitt-Gray studied the
self-selected landings from three different heights. It is
possible that minor changes in techniques may influence the
biomechanical responses. It should be also noted that the
subjects involved in this study were regular and active
university students. whereas in McNitt-Gray's study, elite
and recreational gymnasts were used as the subjects. It was,
demonstrated that the strategies were different between the
two different subject groups ( 18). In the present research the
three self-selected landing strategies were systetnatically
manipulated at three different heights. This is the tirst time
5 18 Official Journal of the American College of Sports Medicine
that all three possible techniques from three heights were
systematically studied and both joint kinematic and kinetic
measurements were presented. In our study. it was clearly
demonstrated that the shift of energy absorption was fro[,,
distal to proximal muscle groups with increased mechanical
demand (i.e. increases in Ianding height). At the lower
height. ankle and knee muscle groups were more effective in
energy dissipation in the stiff Ianding, whereas the knee and
hip extensors were more involved in the soft landing. As the
height and mechanical loading increased. the lower extrem-
ity posture became more tlexed. and the hip extensors be-
came increasinsly involved in energy dissipation. A Second
possible explanation may come from the differences in the
control of techniques in the studies. In the present study, the
three Ianding techniques were self-selected and predefined
in a pretest session using an electrogoniometer for-each
subject in each test condition. The range of maximum knee
flexion using the formula shown earlier was used to monitor
a subject's Ianding technique during the subsequent test
session. Because the joint range of motion for each tech-
nique was different for each subject, this control process
preserved the natural techniques of each subject under the
various test conditions. Even though there is no documented
evidence that a change in natural technique will lead to
changes in joint kinetics during landing, it is logical to
assume that such changes may occur. More research is
warranted_ for this aspect of landing biomechanics.
Main finding in this study was revealed by examining the
relative contribution of each muscle group to the total en-
ergy reduction. Although the absolute eccentric work data
for the knee and hip joints suggested the increases in the
energy absorption with the decreased landing stiffness at the
all three different heights (Table 4), the relative work values
indicated the high and consistent outputs from the knee
extensors. The ankle work was increasingly important in the
stiff landing at the lower heights; however, these percent-
ages decreased with an increase in landing heights and
mechanical demands. This is comprehensible due to the fact
that the ankle plantarflexors exhibit lesser capacity for en-
ergy absorption (I). The hip extensors were.involved min-
imally in STL at the low height and they became more
involved as the mechanical demand increased and the hip
joint posture became more tlexed. due to the massive po-
tential of energy reduction for the muscle group. These
observations were supported by the facts that the relative
contribution of each joint Rob1 to the total change of ROM
demonstrated rather similar fashions (Table I).
Landings involved in many sport activities are repetitive
and violent in nature. This type of stress has long been
associated with overuse injuries (22.23). With increased
mechanical demands, loading to lower extremity joints es-
calates accordingly. It was demonstrated in this study that
the knee joint consistently experienced enormous mechan-
ical demands and high mechanical output in the different
landing conditions. At the greater mechanical demands,
most of energy is absorbed by the massive knee and hip
extensors. Therefore. the eccentric strength and proper
, .~nuscularcontrol of these muscle groups are critical to
integrity of the lower extremity. Finally, it was observed
blltthe interactions of height and technique were significant
for the knee ROM and second peak moment. These inter-
,,[ions may be due to differences in individual landing
and strategy, and further study is warranted.
This study was partially supported by the Orthopedic and Frac-
ture Clinic, Eugene, Oregon. The authors are grateful for the con-
structive comments and recommendations from two anonymous
reviewers.
Address for correspondence: Song-Ning Zhang. Department of
Exercise Science and Sport Management, Biomechanics/Sports
Medicine Laboratory, The University of Tennessee, Knoxville, 1914
Andy Holt Ave., Knoxville, TN 37996; E-mail: szhang@utk.edu.

REFERENCES
I ~LLKANDER. R. M.,and R. F. KER.The architecture of the muscles. 14. HANAVAN. E. P. A mathematical model of the human body. Tech-
In: Miiltiple Miucle Systerns: Biomecl~anicsand Movement Orgn- nical Report (TR-64-102). Aerospace Medical Research Labora-
r~i:ation,J. M. Winters and S. L. Woo (Eds.). New York: Springer- tory. Wright-Patterson Air Force Base, OH, 1964.
Verlag. 1990. pp. 568-577. 15. HUTCH~SON. M. R.. and M. L. IRELAND. Knee injuries in female
2. BOBBERT. M. F., and G. J. V;\N INGENSCHENAU. Coordination in athletes. Sports Med 19:288-302, 1995.
vertical jumping. J. Biomeclr. 2 1:349-262. 1988. 16. JAMES, S. L.,B. T. BATES,and L. R. OSTERNIG. Injuries to runners.
3. COLLINS, K., M. WAGNER. K. PETERSON, and M. STOREY. Ove~sr Am. J. Sports Med. 6:40-49, 1978.
injuries in trirtthletes: a study in 1986 Seafair Triathlon. 17. McNrrr-GRAY. J. L. The influence of impact speed on joint kine-
Am. J. Sports Med. 17:675-680. 1989. matics and impulse characteristics of drop landings. In: Proc ISB.
4 . DEVITA, P.. and W. A. SKELLY. Effect of landing stiffness on joint R. J. Gregor, R. F. Zernicke and W. C. Whiting (Eds.). Los.
kinetics and energies in the lower extremity. Med. Sci. Sports
Angeles: UCLA, 1989.
Exrrc. 24:108-1 15, 1992.
18. MCNIIT-GRAY, J. L. Kinetics of the lower extremities during drop
. . DUFEK, J. S. The effects of landing on lower extremity function.
Doctoral Dissertation. Eugene: University of Oregon. Dept. of landing from three heights. J. Biomech. 26:1037-1046, 1993.
Exercise and Movement Science, Eugene, Oregon. 1988. 19. MOLNAR, T. J., and J. M. FOX.Overuse injuries of the knee in
6. DUFEK,J. S., and B. T. BATES. Models incorporating height, basketball. Clin. Sports Med. 12:349-362. 1993.
distance and landing technique to predict impact forces. In: Pro- 20. PANZER. V. P.. G. A. WOOD,B. T. BATES,and B. R. MASON.Lower
ceedings of XIIth International Cot~gressof Biomechanics, R. J. extremity loads in landings of elite gymnasts. In: Biomechanics
Gregor, R. F. Zirnicke and W. C. Whiting. Los Angela: UCLA. XI-B. Amsterdam: Free University Press. 1988, pp. 727-735.
ISSN: 002 19290, 1989. 21. PRILUTSKY. B. I., and V. M. ZATSIORSKY. Tendon action of two-
7. DUFEK, J. S., and B. T. BATES.The evaluation and prediction of joint muscles: transfer of mechanical energy between joints during
impact Force during landings. Med. Sci. Sports Exerc. 22:370-376, jumping. landing, and running. J. Biomech. 27:25-34, 1994.
1990. 22. RADIN.E. L., and P. I. L. Response of joints to impact loading. I.
1. DUFEK, J. S., and B. T. BATES.Regression models for predicting Arthritis Rheum. 14:356-362. 1971.
impact forces and knee joint moments and power during landings. 23. RADIN. E. L., B. M. MARTIN, D. B. BURR.B. CATERSON, R. D. s o d ,
In: Proc. CSB, 1990, pp. 55-56. and C. GOODWIN. Effects of mechanical loading on the tissues of
9. DUFEK.J. S., and B. T. BATES.Lower extremity performance the rabbit knee. J. Orthop. Rcs. 2:221-234, 1984.
models for landing. Hiun. Mov. Sci. 1 1 999-3 18. 1992. 24. R E N ~ U P. M A., Knee pain in tennis players. Clin. Sports Med.
10. ERIKSSON, E.. and R. J. JOHNSON. The etiology of downhill ski 14: 163-175, 1995.
injuries. Exerc. Sport Sci. Rev. 8:l-17, 1980. , X.KAELIN,
25. S r ~ c o mA., and E.S ~ m s rThe
. impact in landing after
11. FERRET~I, A., G. PUDDU, P. P. MARIANI, and M. NERI.Jumper's
a volleyball block. In: Biomechanics 10-B. Champaign, IL.: Hu-
knee: an epidemiological study of volleyball players. Pliysician
Sportsmed. 12:97-106, 1984. man Kinetics, 1988. pp. 694-700.
12. GERBERICH. S. G., S. LUHMANN, C. FINKK, J. D. PRIEST,and B. J. 26. WINTER, D. A. Biomeclaanic~and Motor Control of Human Move-
BEARD. Analysis of severe injuries associated with volleyball ac- menr. 2nd Ed. New York: Wiley, 1990, pp. 75-102.
tivities. Physician Sporrsmed. 1275-80. 1987. 27. YAMAGUCHI, G. T., A. G.-U. SAWA, D. W. MORAN, M. J. FESSLER,
13. GRAY.J.. J. E. TAUN~ON. D. C. MCKENLIE, D. B. CLEMENT, J. P, and J. M. WINTERS.A survey of human musculotendon actuator
MCCONKEY, and R. G. DAVIDSON. A survey of injuries to the parameters. In: Mctltiple Muscle Systems: Biomechanics and
anterior cruciate ligament of the knee in female basketball players. Movement Organization, J. M . Winters and S. L. Woo (Eds.). New
Int. J. Sports Med. 6:3 14-3 16, 1985. York: Springer-Verlag. 1989, pp. 717-773.

LANDING BIOMECHANICS Medicine & Science in Sports & exercise^ 8 19