Hip Moments During Level Walking, Stair Climbing, and Exercise in Individuals Aged 55 Years or Older.

Key Words: Gait analysis, Hip fracture, Joint moments of force, Osteoporosis, Rate of change in moments.

Exercise is an intervention that may decrease the likelihood of hip fractures in adults over 65 years of age by reducing bone loss.[1-4] This recommendation is based on studies demonstrating the detrimental effects of immobilization and weightlessness on bone density and demonstrating the positive relationship between high levels of physical activity in childhood and bone mass at skeletal maturity.[5-9] There is evidence that exercise is effective in increasing bone density or at least in reducing the rate of bone loss in women who are postmenopausal[10-13] and in increasing bone density in young women.[14] However, there is only one report of improvement in bone mass of the proximal femur that occurred as a result of exercise.[15]

In studies in which exercise had no beneficial effect,[16-18] the authors suggested that exercise intensity may have been insufficient to cause a change in the mass of the proximal femur. The amount and rate of application of mechanical load and the diversity of stresses applied are more important in influencing bone remodeling than are the aerobic intensity of the exercise or the number of load cycles.[19] Using animal models, Lanyon et al[20] hypothesized that bone remodeling is influenced by a feedback mechanism that operates to maintain bone strain levels at an optimum value. According to this hypothesis, adaptive bone remodeling adjusts the amount of tissue present according to an increase or a decrease in the bone strain levels.

Rubin and Lanyon[21] showed that bone remodeling is sensitive to both strain distribution and strain magnitude. Their study indicated that the intensity, diversity, and rate of rise of the mechanical stress are more important than is the number of load cycles.[19,20] Similarly, Burr et al[22] demonstrated that loads of 1.5 times body weight applied for 20 minutes per day for 3 weeks were sufficient to promote bone formation in the vertebral columns of rabbits. Burr and colleagues also contend that exercise involving repetitive loading of sufficient intensity of the lower limbs could be used to prevent bone loss in the spine.

Several researchers have attempted to quantify the mechanical effect of exercise on the proximal femur.[2,10,23] Bassey and Ramsdale[2] used a force plate to measure the peak impact forces during a heel-drop exercise that involved raising the body weight onto the toes and then letting the heels drop to the floor while keeping the knees locked and the hips extended. Similarly, Grove and Londeree[10] categorized exercises as low-or high-intensity activities on the basis of peak forces obtained with a force platform. Woodward and Cunningham,[23] on the basis of research with an accelerometer, reported a higher rate of change of acceleration at the ankle during running and ascending and descending stairs than during walking and cycling. Although these studies described measurement systems for the quantitative assessment of exercise, neither acceleration nor peak forces predict the magnitude of the mechanical work done at the hip joint during a particular activity. Net moments of force at the hip are needed to indicate the mechanical load on the proximal femur resulting from various exercises. Joint moments of force are the net result of all internal forces acting at the joint and include the moments attributable to muscles, bones, ligaments, and other soft tissues around the joint.

The purpose of our study was to investigate various exercises to determine which exercises result in the greatest moments at the hip in individuals aged 55 years or older. Knowledge of the net mechanical effect of walking, stair climbing, and other exercises at the hip joint would, theoretically, allow the development of an optimal exercise program designed to maintain or increase femoral bone mass.

Method

Subjects

Data were collected from 17 Caucasian men and 13 Caucasian women who were 55 years of age or older and had no identified musculoskeletal or neurological impairments. Subjects were recruited through newspaper advertising. The mean age of the subjects was 65.4 years (SD=6.02, range=55-75) (for men: [bar]X=66.6, SD=5.6, range=57-75; for women: [bar]X=63.7, 8D=6.4, range=55-73). The mean weight of the subjects was 80.5 kg (SD=16.2, range=49.5-114) (for men: [bar]X=87.6, SD=13.5, range=71-114; for women: [bar]X-71.2, SD=15.0, range=49.5-95). The mean height of the subjects was 170.8 cm (SD=10.5, range= 153-187) (for men: [bar]X=177.9, SD=6.1, range=162-187; for women: [bar]X=161.3, SD=6.8, range=153-173).

Measurement System

Hip internal moments during level walking, stair climbing, and exercises were obtained with the QGAIT system.(*)[24,25] The QGAIT system is a software package that works in conjunction with the Optotrak motion system,([dagger]) a force plate, standardized radiographs (which provide information on the location of the joint centers), and anthropometric data.

Infrared-emitting diodes (IREDs) were placed over the greater trochanter, the lateral femoral epicondyle, the head of the fibula, and the lateral malleolus. Two probes, which projected anteriorly, were attached to each subject's thigh and shank, and an IRED was attached to these endpoints, providing 3 markers per limb segment. Three IREDs attached to a plastic anchor were placed at the level of fifth lumbar vertebra. A mechanical on-off footswitch placed under the right heel provided inibrmation regarding foot contact during gait and stair ascent and descent to define a gait cycle for normalization of the data.

Vertical and shear ground reaction forces were measured with an AMTI force platform.([double dagger]) A 16-channel analog-to-digital board integrated with the Optotrak system allowed simultaneous collection of motion, force plate, and footswitch data. Force and motion data were collected at a frequency of 50 Hz.

Standardized radiographs[26,27] were used to calculate the discrepancy between the surface location of the IREDs and the positions of the hip and knee joint centers. Lead shots were placed over the bony IRED landmarks (greater trochanter, lateral femoral epicondyle, and head of the fibula). These lead shots were then located on radiographs; the discrepancy between the external bony landmark and the actual hip and knee joint centers was estimated, and a correction factor was calculated. The location of the ankle joint center was estimated as the midpoint of the distance between the lateral and medial malleoli, which was measured with a caliper.

The anthropometric data collected included each subject's height, weight, lengths of lower limb from greater trochanter and tibial plateau to the floor, thigh and calf circumferences, and shoe weight. These measurements are necessary for the calculation of thigh, shank, and foot center-of-mass locations; segment masses; and inertial properties.

Kinematic, kinetic, anthropometric, and joint center data were integrated in order to calculate net moments (in newton-meters) during level walking, stair climbing, and exercise. Hip moments were obtained by use of inverse dynamics with a link segment model. In the inverse-dynamics approach, the segments are assumed to be rigid bodies with hinge joints between them. In our model, the foot was considered part of the shank; therefore, our model was composed of shank and thigh segments only.[25] The moments were calculated first at the proximal end of the shank and then at the proximal end of the thigh. Therefore, moments of force were calculated first at the knee and then at the hip. The moments obtained during each activity were normalized by body weight, resulting in measurements expressed in newton-meters per kilogram. The rate of change in moments was obtained in all 3 planes by use of a program designed to locate the maximum slope during each exercise cycle and was reported in newton-meters per kilogram per second.

The QGAIT system was previously used to obtain measurements of knee moments and forces, and measurements of these gait variables were found to be accurate and reproducible.[24,25] The test-retest reliability of measurements of 3-dimensional moments of force at the hip during gait was determined in our laboratory for 10 subjects (5 men, 5 women) who had a mean age of 27 years. Testing involved 2 visits to the laboratory over a 2- week period. The hip moments demonstrated excellent repeatability between visits, with the 95% confidence intervals of the maximal point difference between curves for all 6 moments crossing zero in all planes.

The exercises we examined were those likely to generate high net moments and rate of change in moments at the hip, suitable for older individuals, and measurable by the system. The exercises included weight-bearing activities, such as left hip flexion, left hip flexion with a 1.13-kg weight around the left ankle, left hip extension, and left hip abduction while in single-limb stance on the right test leg. In addition, ascending stairs and descending stairs, basic or aerobic steps, lunge, and knee bend were investigated (Tab. 1). The non-weight-bearing activities included right test hip flexion, right test hip flexion with a 1.13-kg weight around the right ankle, right test hip extension, and right test hip abduction while subjects stood on the left leg (Tab. 1). Figures 1 through 3 demonstrate a subject performing flexion of the right test hip, basic step, and flexion of the left hip with weight, respectively. All subjects wore running shoes for all tests.

Table 1.

Description of the Exercises Investigated in the Study

Exercise                      Subject's Position and Action

1. Level walking              Walking over a walkway that has
                                a force plate embedded on it.

2. Ascending stairs           A stair's width from the force
                                plate. The height of the
                                staircase was 21.5 cm.

3. Descending stairs          On the step above the force plate
                                about a stair's width away
                                from the force plate,

4. Basic step                 Standing with both feet on
                                top of the stepper
                                (16 cm), which was located
                                in front of the force
                                plate. Step backward and
                                down with the left
                                foot adjacent to the force plate
                                and then bring
                                the right foot onto the plate.

5. Knee bend                  Standing beside the force plate.
                                Bring the right foot onto the
                                force plate, bend both right and
                                left knees, straighten knees,
                                and return to the initial
                                position.

6. Forward lunge              Standing with both feet 20
                                cm from the force plate. Step
                                forward with the right foot onto
                                the force plate, taking the weight
                                onto the front leg
                                with the right knee bent, and
                                then return to the start position.

7. Flexion of                 Standing up straight with both
    the left hip                feet adjacent to the
                                force plate. Move the right
                                test leg onto the force plate and
                                then  raise the left leg straight
                                forward up and down, keeping
                                the knee straight.

8. Abduction of               Standing up straight with both feet
    the left hip                adjacent to the force plate. Move
                                right test leg onto the plate
                                and then raise left leg sideways
                                up and down, keeping the knee
                                straight.

9. Extension of               Standing up straight with both
    the left hip                feet adjacent to the
                                force plate, Move the right
                                test leg onto the
                                force plate and then bring the
                                left leg straight
                                backward and down, keeping
                                the knee straight.

10. Flexion of the            Standing up straight with both
     left hip with weight       feet adjacent to the
                                force plate. Move the right
                                test leg onto the
                                force plate and then bring
                                the left leg straight
                                forward up and down,
                                keeping the knee
                                straight. There is a 1.13-kg
                                nylon band around
                                the ankle.

11. Flexion of the            Standing up straight. Move the
     right test hip             right leg straight
                                forward up and down, keeping
                                the knee straight.

12. Abduction of the          Standing up straight. Move the
     right test hip             right leg sideways
                                up and down, keeping the
                                knee straight,

13. Extension of the          Standing up straight. Move the
     right test hip             right leg straight
                                backward up and down, keeping
                                the knee straight,

14. Flexion of the            Standing up straight. Move the
     right test hip             right leg straight
     with weight                forward up and down,
                                keeping the knee
                                straight. There is a 1.13-kg
                                nylon band around the ankle.

Exercise                      Exercise Cycle

1. Level walking              From foot contact of the test leg
                                on the force plate and ending with
                                the next foot contact of the test
                                leg on the walkway.

2. Ascending stairs           The step of interest begins with
                                toe-off of the right
                                leg from the ground,
                                continues through foot
                                contact with the force plate,
                                and ends with toe-
                                off of the right leg from
                                the force plate.

3. Descending stairs          The step of interest begins
                                with right foot contact
                                on the force plate, continues
                                through toe-off of
                                the right leg from the force
                                plate, and ends with
                                toe contact on the ground.

4. Basic step                 From right foot contact on
                                the force plate and
                                continuing through toe-off of
                                the right test leg
                                from the plate.

5. Knee bend                  The cycle consisted of the period
                                during which the subject's right
                                test leg was on the force plate.

6. Forward lunge              The cycle consisted of the
                                period during which the
                                subject's right test leg was
                                on the force plate.

7. Flexion of                 The cycle consisted of the
    the left hip                period during which the
                                right test leg was on
                                the force plate.

8. Abduction of               The cycle consisted of the
    the left hip                period during which the
                                right test leg was on
                                the force plate.

9. Extension of               The cycle consisted of the
    the left hip               period during which the
                               right test leg was on the
                               force plate.

10. Flexion of the            The cycle consisted of the
     left hip with weight       period during which the
                                right test leg was on
                                the force plate.

11. Flexion of the            The cycle extended from the
     right test hip             neutral position to
                                maximum hip flexion to the
                                neutral position again.

12. Abduction of the          The cycle extended from the
     right test hip             neutral position to maximum hip
                                abduction to the neutral position
                                again.

13. Extension of the          The cycle extended from the
     right test hip             neutral position to
                                maximum hip extension to the
                                neutral position again.

14. Flexion of the            The cycle extended from the
     right test hip             neutral position to maximum hip
     with weight                flexion to the neutral position
                                again.

Procedure

All subjects signed a consent form prior to participating in the study. Radiographs of the hip and knee were obtained first. The 3 anatomical landmarks (greater trochanter, lateral femoral epicondyle, and head of the fibula) were identified on the right leg and marked with an erasable marker, and a small lead shot was placed over each landmark for radiographic purposes. In the motion laboratory, IREDs were attached as previously described. A footswitch was placed under the right heel for gait and stair-climbing data collection.

Data from the right lower limb were collected during all activities. Level walking was performed by all 30 subjects to provide the baseline measure used for comparison with the other exercises. Data were collected for 24 subjects during ascending stairs. Descending stairs and the remaining exercises were randomized so that each subject performed 3 or 4 of the exercises under investigation. Subjects were provided with an opportunity to practice each activity prior to recording and were provided with rest periods as necessary between activities. Data for 5 trials were collected for each activity, and the average of the 5 trials was used in data analysis. One visit of approximately 2 hours was required for the data collection.

Data Analysis

A 2-way analysis of variance (ANOVA) (2 exercises X 3 planes) for repeated measures on all factors was carried out with the maximum peak moment and rate of change in moments obtained at the hip during the activities under investigation. Each exercise was compared with level walking. When the F ratio was significant, a paired-sample t test was conducted to determine where the differences between level walking and each of the exercises occurred. A t-test power analysis was performed when no significant difference was found. All statistical tests were considered significant at the .05 level.

Results

Peak Internal Moments

The overall results obtained for the hip peak internal moments during level walking and other exercises are shown in Table 2. The ANOVA test comparing level walking with each of the exercises demonstrated a difference in moments between planes in each case. Differences for exercise were found when level walking was compared with ascending stairs, knee bend, lunge, basic step, and abduction of the left hip. The interactions between exercise and plane for comparisons between level walking and all the other exercises were significant, except for lunge. The paired-sample t test between level walking and each of the exercises was performed to compare means in each plane. All comparisons showed a t-test power above 75%.

Table 2.

Comparison of the Mean Maximum Hip Peak Moments During Level Walking, Stair Climbing, and Exercise(a)

                Plane

                Frontal               Sagittal

Exercise   N    MxAbd     MxAdd       MxFlx      MxExt

LW         30   0.81      -0.17        0.66      -0.89

LW         24   0.82      -0.18        0.68      -0.94
AS              0.77      -0.07(b)     0.28(b)   -1.00

LW          8   0.91      -0.18        0.78      -1.03
DS              0.96      -0.04(b)     0.20(b)   -0.50(b)

LW          9   0.83      -0.26        0.65      -1.04
BS              0.76      -0.01(b)     0.35(b)   -0.43(b)

LW          9   0.87      -0.18        0.67      -1.09
KB              0.24(b)   -0.18        0.09(b)   -0.76(b)

LW          7   0.76      -0.22        0.65      -0.86
LU              0.35(b)   -0.11        0.29      -0.99

LW          7   0.80      -0.12        0.55      -0.84
FL              0.85      -0.09        0.18(b)   -0.47(b)

LW          7   0.77      -0.15        0.64      -0.88
FLW             0.79      -0.12        0.18(b)   -0.52(b)

LW          8   0.86      -0.15        0.78      -0.96
AL              0.83      -0.14        0.04(b)   -0.60(b)

LW          8   0.84      -0.21        0.64      -1.07
EL              0.75      -0.08(b)     0.1(b)    -0.60(b)

LW          8   0.81      -0.13        0.63      -0.84
FR              0.12(b)   -0.01(b)     0.40(b)   -0.04(b)

LW          7   0.84      -0.19        0.60      -0.92
FRW             0.20(b)   -0.003(b)    0.47      -0.03(b)

LW          9   0.78      -0.13        0.60      -0.81
AR              0.31(b)   -0.02(b)     0.16(b)   -0.03(b)

LW          6   0.85      -0.20        0.68      -0.90
ER              0.20(b)    0.07(b)    -0.14(b)   -0.46

           Plane

           Transverse

Exercise   MxERot       MxIRot

LW          0.08        -0.10

LW          0.08        -0.11
AS          0.10        -0.21(b)

LW          0.09        -0.12
DS          0.09        -0.14

LW          0.09        -0.12
BS          0.03(b)     -0.04(b)

LW          0.09        -0.13
KB          0.06        -0.07(b)

LW          0.07        -0.09
LU          0.11        -0.06

LW          0.07        -0.09
FL          0.03(b)     -0.03(b)

LW          0.07        -0.08
FLW         0.03(b)     -0.03(b)

LW          0.09        -0.11
AL          0.03(b)     -0.04(b)

LW          0.09        -0.12
EL          0.03(b)     -0.04(b)

LW          0.08        -0.09
FR          0.002(b)    -0.002(b)

LW          0.08        -0.10
FRW         0.004(b)    -0.005(b)

LW          0.07        -0.09
AR          0.006(b)    -0.007(b)

LW          0.10        -0.12
ER         -0.009(b)    -0.01(b)

(a) MXAbd=maximum abductor moment, MxAdd=maximum adductor moment, MXFlx=maximum flexor moment, MXExt=maximum extensor moment, MXERot=maximum external rotator moment, MXIRot=maximum internal rotator moment, LW=level walking, AS=ascending stairs, DS=descending stairs, BS=basic step, KB=knee bend, LU=lunge, FL=flexion of the left hip, FLW=flexion of the left test leg with weight, AL=abduction of the left leg, EL=extension of the left leg, FR=flexion of the right test leg, FRW=flexion of the right lest leg with weight, AR=abduction of the right test leg, ER=extension of the right test leg. Values are reported in newton-meters per kilogram per second.

(b) Significantly different from value for LW at P [is less than or equal to] .05, as determined by ANOVA and paired sample t test.

Frontal plane. The maximum mean peak internal abductor moment was obtained during descending stairs and reached 0.96 N [multiplied by] m/kg; the value obtained during level walking was 0.91 N [multiplied by] m/kg. Flexion of the left hip and flexion of the left hip with weight showed internal abductor moments of the right test leg that were similar to those obtained during level walking (0.85 versus 0.80 N [multiplied by] m/kg and 0.79 versus 0.77 [multiplied by] N [multiplied by] m/kg, respectively). Similarly, the moments generated during ascending stairs, basic step, and abduction and extension of the left hip were not different from those obtained during level walking. The knee bend, lunge, and non-weight-bearing test leg exercises (abduction, extension, and flexion of the right test hip and flexion of the right test hip with weight) showed peak internal abductor moments lower than the mean peak value obtained during level walking.

The magnitude of the internal adductor moment obtained during level walking was higher than those obtained during most of the exercises. The magnitudes of the internal adductor moments during knee bend, lunge, flexion and abduction of the left hip, and flexion of the left hip with weight were not different from those obtained during level walking. The remainder of the exercises generated hip internal adductor moments that were lower than those obtained during level walking.

Sagittal plane. In the sagittal plane, none of the exercises resulted in peak internal flexor moments higher than those obtained during level walking. A comparison of the internal flexor moments between lunge and level walking (0.29 versus 0.65 N [multiplied by] m/kg) and between flexion of the right test hip with weight and level walking (0.47 versus 0.60 N [multiplied by] m/kg) resulted in no difference. The remainder of the exercises showed peak internal flexor moments lower than those obtained during level walking.

The maximum internal extensor moments obtained during ascending stairs (-1.0 N [multiplied by] m/kg), lunge (-0.99 N [multiplied by] m/kg), and extension of the right test hip (-0.46 N [multiplied by] m/kg) were not different from those obtained during level walking. All the other exercises generated peak internal extensor moments that were lower than those obtained during level walking.

Transverse plane. In the transverse plane, the external rotation moments obtained during lunge and ascending stairs were not different from those obtained during level walking (0.11 versus 0.07 N [multiplied by] m/kg and 0.10 versus 0.08 N [multiplied by] m/kg, respectively). Similarly, descending stairs and knee bend generated hip external rotation moments comparable to those obtained during level walking. All the other exercises generated hip external rotation moments that were lower than those obtained during level walking.

The peak internal rotation moment obtained during ascending stairs was higher than that obtained during level walking (-0.21 versus -0.11 N [multiplied by] m/kg). Descending stairs and lunge generated peak internal rotation moments that did not differ from those obtained during level walking. All the other exercises generated peak moments lower than those obtained during level walking.

Rate of Change in Moments

The mean rates of change in moments obtained during level walking and all the other exercises are shown in Table 3. The results of the ANOVA for the main effect of plane and for the interactions between exercise and plane were all significant.

Table 3.

Comparison of the Mean Maximal Rates of Change in Moments at the Hip During Level Walking, Stair Climbing, and Exercise(a)

                Plane

Exercise   N    Frontal     Sagittal    Transverse

LW         30    0.083      -0.161      -0.017

LW         24    0.096      -0.166      -0.017
AS              -0.060(b)   -0.090(b)   -0.020

LW          8    0.067      -0.179      -0.017
DS               0.111      -0.081(b)   -0.009

LW          9    0.062      -0.177      -0.017
BS               0.067      -0.008(b)   -0.002(b)

LW          9    0.100      -0.174      -0.017
KB               0.017      -0.031(b)    0.003(b)

LW          7    0.084      -0.171      -0.022
LU              -0.023      -0.074      -0.005(b)

LW          7    0.151      -0.166      -0.017
FL               0.033(b)   -0.006(b)   -0.001(b)

LW          7    0.152      -0.18       -0.019
FLW              0.016(b)   -0.018(b)    0.002(b)

LW          8    0.101      -0.18       -0.016
AL               0.026      -0.014(b)   -0.000(b)

LW          8    0.058      -0.158      -0.015
EL               0.022      -0.008(b)   -0.002(b)

LW          8    0.077      -0.158      -0.016
FR               0.001       0.003(b)   -0.000(b)

LW          7    0.062      -0.151      -0.018
FRW             -0.002       0.001(b)   -0.000(b)

LW          9    0.101      -0.175       0.101
AR               0.005(b)    0.002(b)   -0.000(b)

LW          6   -0.033      -0.127      -0.012
ER               0.006      -0.006(b)   -0.001(b)

(a) LW=level walking, AS=ascending stairs, DS=descending stairs, BS=basic step, KB=knee bend, LU=lunge, FL=flexion of the left hip, FLW=flexion of the left test leg with weight, AL=abduction of the left leg, EL=extension of the left leg, FR=flexion of the right test leg, FRW=flexion of the right test leg with weight, AR=abduction of the right test leg, ER=extension of the right test leg. Values are reported in newton-meters per kilogram per second.

(b) Significantly different from value for LW at P [is less than or equal to] .05, as determined by ANOVA and paired sample t test.

In the frontal plane, descending stairs had a rate of change in moments comparable to the mean rate of change obtained during level walking (0.111 versus 0.067 N [multiplied by] m/kg/s). Similarly, basic step, knee bend, lunge, flexion of the right test hip with weight, flexion and extension of the right test hip, and abduction and extension of the left hip had rates of change in moments that did not differ from those obtained during level walking. All the other exercises had rates of change in moments lower than those obtained during level walking.

In the sagittal plane, none of the rates of change in moments during the exercises were higher than those obtained during level walking. The rate of change in moment with lunge did not differ from those obtained during level walking. All the other exercises generated rates of change in moments lower than those obtained during level walking.

In the transverse plane, the rates of change in moments during ascending stairs and descending stairs did not differ from those obtained during level walking. All the other comparisons resulted in rates of change in moments lower than those obtained during level walking.

Discussion

Frontal Plane

The magnitudes of the internal abductor moments in several of the exercises (ascending and descending stairs, flexion of the left hip with and without a weight, abduction and extension of the left hip, and basic step) were comparable to those obtained during level walking. The rates of change in moments during descending stairs, abduction and extension of the left hip, and basic step also did not differ from those obtained during level walking. Therefore, in the frontal plane, all of these exercises could be recommended to promote diversity and high loads at the hip joint.

The internal abductor moment during level walking is attributable primarily to the medial shift of the body's center of mass as the weight is transferred to the stance limb. The abductors of the stance limb create a moment that limits the drop of the contralateral pelvis.[28] During descending stairs, when the foot strikes the step below, the abductors also have to generate moments with enough magnitude to sustain the body weight as it is transferred to the stance limb. The impact of the braking forces as the body moves down during descending stairs is probably comparable to or greater than the impact at heel contact during level walking, creating similar joint reaction forces and consequently moments.

Flexion of the left hip and flexion of the left hip with weight had almost no impact. The moments generated during these exercises, however, were similar to those generated during level walking. To sustain the body weight as the opposite leg goes into flexion apparently requires a level of activity comparable to that required of the same muscles during the stance phase of the gait cycle. Similarly, ascending stairs and abduction and extension of the left hip resulted in internal abductor moments and rates of change in moments that did not differ from those obtained during level walking. Therefore, these exercises may be an alternative to level walking in a program designed to promote high loads at the hip.

The magnitude of the Internal adductor moment obtained in the frontal plane during level walking was, on average, very low and comparable to results reported in the literature.[29,30] The maximum peak internal adductor moment occurred during the swing phase of the gait cycle and was similar in magnitude to the peak adductor moment generated during knee bend. The functions of the adductors during the swing phase of gait are to maintain the body near the midline and to help the flexor muscles flex the hip.[31,32] Knee bend involves greater hip flexion range than does level walking. Thus, it is likely that the demand placed on the adductors during knee bend would be higher to assist the flexors during knee bend.

Similarly, lunge, flexion of the left hip with and without a weight, and abduction of the left hip also generated internal adductor moments that did not differ from those obtained during level walking. In addition, the rates of change in moments during knee bend, lunge, and abduction of the left hip were not different from those obtained during level walking. Thus, these exercises could be used in addition to level walking as components of a program designed to create high loads at the hip in the frontal plane.

Sagittal Plane

According to Andriacchi and Mikosz,[33] the highest external moments during most activities of daily living occur in the sagittal plane, in the direction tending to flex the joints. Thus, the demand is placed on the extensor antigravity muscles of the joint to counterbalance the external flexor moments.[33,34] The internal extensor moment obtained during level walking in the current study reached -0.89 N [multiplied by] m/kg. This magnitude is in agreement with those in other studies in which sagittal moments at the hip were measured.[29,30]

The activities that generated the highest internal extensor moments in the sagittal plane were ascending stairs and lunge, although these values were not higher than those obtained during level walking. As the intensity and physical demands of a task increase, the external moments and the internal forces increase proportionally.

The fastest rise time in moments in the sagittal plane was observed during level walking. Lunge showed a rate of change in moment of a magnitude similar to that obtained during level walking. Therefore, in the sagittal plane, lunge and ascending stairs would be appropriate for generating high internal extensor moments at the hip.

Transverse Plane

The magnitude of the internal rotator moment during ascending stairs was higher than that obtained during level walking. Ascending stairs also generated a rate of change in moment comparable to that obtained during level walking. Therefore, ascending stairs would appear to be the best activity for promoting high loads with a fast rise time at the hip in the transverse plane.

Descending stairs and lunge resulted in hip internal rotation moments comparable to those obtained during level walking. Similarly, knee bend, lunge, and ascending and descending stairs generated hip external rotation moments comparable to those obtained during level walking. Consequently, these activities may be recommended as alternatives to level walking for enhancing mechanical loads at the hip in the transverse plane.

In general, the non-weight-bearing exercises (flexion of the right test hip, flexion of the right test hip with weight, abduction of the right test hip, and extension of the right test hip) generated lower moments of force at the hip than did the weight-bearing activities in almost all planes. Flexion of the right test hip with weight and extension of the right test hip generated moments that were not different from those obtained during level walking in the sagittal plane only. The remainder of these exercises (abduction of the right test hip and flexion of the right test hip) generated moments that were lower than those generated during level walking in every plane. In these types of exercises, the external moments are the product of the acceleration and the weight of the leg. External forces involved in the system are inertia, gravity, and forces attributable to the distal segment. Thus, it is not surprising that the moments generated during these exercises were lower than those generated during the weight-bearing test leg exercises. The weight placed around the ankle during flexion of the right test hip was an attempt to increase the moments generated at the hip. As the mass of the segment increases, the applied forces have to increase to change the state of motion of the segment. A weight of 1.13 kg, however, was not sufficient to promote high moments at the hip. How much more weight would need to be added to the lower extremity to generate peak moments higher than those obtained during level walking in each plane should be determined in future research.

Burr et al[22] demonstrated that repetitive loading of up to 1.5 times body weight of the lower limb initiated bone remodeling in the vertebral columns of rabbits. On the basis of this result, Grove and Londeree[10] hypothesized that activities that generate impact forces greater than 1.5 times body weight, such as jumping jacks, knee to elbow with jump, and running in place, would be required to maintain or increase the bone mass density of the proximal femur in women who are postmenopausal. Activities with an impact less than 1.5 times body weight, such as slow and fast walking, were thought to be insufficient for bone mass maintenance. On the basis of this information, none of the exercises investigated in this study would be recommended, because the moments generated in every plane were lower than 1.5 times body weight. Ground reaction forces, however, were the outcome measure reported in the study by Grove and Londeree.[10] Ground reaction forces represent the net vertical and shear forces acting between the foot and the ground (force plate) and are equal in magnitude to the force that the body applies to the ground through the foot.[31,35] These forces are the sum of the mass x acceleration products of all body segments while the foot is in contact with the force plate.[35] These forces represent the total body reaction forces, rather than the forces at the hip joint specifically. In our opinion, these ground forces would not yield information regarding any treatment decision related to a specific joint.[35] The moments generated at the hip during the high-impact exercises described by Grove and Londeree[10] could be lower than 1.5 times body weight. Therefore, the moments of force generated at the hip during each activity should be investigated prior to any exercise recommendation.

Woodward and Cunningham,[23] using an accelerometer attached to the ankle, showed that the rates of change in acceleration at the ankle during ascending and descending stairs were higher than those obtained during level walking. In our study, the rates of change in moments at the hip during descending stairs (frontal plane) and ascending stairs (frontal and transverse planes) were comparable to those obtained during level walking. A comparison between these 2 studies is difficult because both the measurement systems and the outcome measures used in the studies were different. In addition, the rates of change of acceleration reported by Woodward and Cunningham[23] do not reflect the rates of change in forces. The measurement of moment of force during various exercises provides an indication of the relative force acting at a joint.[34] Therefore, we suggest that the rate of change in moments is a reasonable measure of the rate of change of force and hence strains.

Conclusion

Our findings indicate that none of the exercises investigated resulted in higher moments and rates of change in moments at the hip than those recorded during level walking in all 3 planes. Most of the exercises examined generated moments with magnitudes and rates of change in moments that were not different from those obtained during level walking in one plane only. Co-contraction of opposing muscle groups may occur during some exercises, for instance, during single-limb stance, generating higher moments at the hip. The QGAIT system, however, does not take co-contraction into account, possibly resulting in the measurement of lower moments than actually exist.

The appropriate exercise prescription for the prevention of bone loss in the proximal femur has not been well-established. Studies of the mechanical forces at the hip during different exercises provide information about the potential of the activities to enhance bone density. Repetitive loads applied in habitual situations are not believed to be sufficient to promote bone formation.[19] The osteogenic response appears to adapt to the amount and direction of the mechanical load applied. Therefore, level walking as a normal everyday activity alone may not alter the osteogenic response. Walking with the addition of weights around the waist or on the trunk might increase the load at the hip joint and possibly induce an osteogenic response. We recommend that an exercise program designed to minimize or reverse the reduction in bone mass occurring with osteoporosis should incorporate exercises that generate moments higher than or comparable to those generated by level walking. The incorporation of these exercises could add both intensity and diversity of stress, which are important for maximal increase in bone formation.[19,21,36] Further research, however, is necessary to determine the effectiveness of exercise programs in minimizing bone loss in the proximal femur and in preventing fractures in older adults.

Acknowledgments

We thank Scott Kirkwood for technical support and STATLAB, Queen's University, for statistical advice.

(*) Clinical Mechanics Group, Queen's University, Kingston, Ontario, Canada K7L 3N6.

([dagger]) Northern Digital Inc, 403 Albert St, Waterloo, Ontario, Canada N2L 3V2.

([double dagger]) Advanced Medical Technology Inc, 176 Wahham St, Watertown, MA 02172.

References

[1] Orwoll ES, Bauer DC, Vogt TM, Fox KM. Axial bone mass in older women. Ann Intern Med. 1996;124:187-196.

[2] Bassey EJ, Ramsdale SJ. Weight-bearing exercise and ground reaction forces: a 12-month randomized controlled trial of effects on bone mineral density in healthy postmenopausal women. Bone. 1995; 16:469-476.

[3] Drinkwater BL. Exercise in the prevention of osteoporosis. Osteoporos Int. 1993;13(suppl 1):169-171.

[4] Gutin B, Kasper MJ. Can vigorous exercise play a role in osteoporosis prevention? A review. Osteoporos Int. 1992;2:55-69.

[5] Sinaki M. Exercise and osteoporosis. Arch Phys Med Rehabil. 1989;70: 220-229.

[6] Sinaki M, Offord KP. Physical activity in postmenopausal women: effect on back muscle strength and bone mineral density of the spine. Arch Phys Med Rehabil. 1988;69:277-280.

[7] Dempster DW, Lindsay R. Pathogenesis of osteoporosis. Lancet. 1993;341:797-801.

[8] Pocock N, Eisman J, Gwinn T, et al. Muscle strength, physical fitness, and weight but not age predict femoral neck bone mass. J Bone Miner Res. 1989;4:441-448.

[9] Cummings SR, Kelsey JL, Nevitt MC, O'Dowd KJ. Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol Rev. 1991 ;7:178 -208.

[10] Grove KA, Londeree BR. Bone density in postmenopausal women: high impact vs low impact exercise. Med Sci Sports Exerc. 1992;24: 1190-1194.

[11] Hatori M, Hasegawa A, Adachi H, et al. The effects of walking at the anaerobic threshold level on vertebral bone loss in postmenopausal women. Calcif Tissue Int. 1993;52:411-414.

[12] Dalsky GP, Stocke KS, Ehsani AA, et al. Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann Inter, Med. 1988;108:824-828.

[13] Pruitt LA, Jackson RD, Bartels RL, Lehnhard HJ. Weight-training effects on bone mineral density in early postmenopausal women. J Bone Miner Res. 1992 ;7:179-185.

[14] Snow-Hatter C, Bouxsein ML, Lewis BT, et al. Effects of resistance and endurance exercise on bone mineral status of young women: a randomized exercise intervention trial. J Bone Miner Res. 1992;7: 761-769.

[15] Nelson ME, Fiatarone MA, Morganti CM, et al. Effects of high-intensity strength training on multiple risk factors for osteoporotic fractures: a randomized controlled trial. JAMA. 1994;272:1909-1914.

[16] Bornor JA, Dilworth BB, Sullivan KM. Exercise and osteoporosis: a critique of the literature. Physiotherapy Canada. 1988;40:146-155.

[17] Smith EL, Gilligan C. Physical activity effects on bone metabolism. Calcif Tissue Int. 1991;49 (suppl):S50-S54.

[18] Block JE, Smith R, Friedlander A, Genant HK. Preventing osteoporosis with exercise: a review with emphasis on methodology. Med Hypotheses. 1989;30:9-19.

[19] Lanyon LE. Control of bone architecture by functional load bearing. J Bone Miner Res. 1992;7 (suppl 2):S369-S375.

[20] Lanyon LE, Goodship AE, Pye CJ, MacFie JH. Mechanically adaptive bone remodelling. J Biomech. 1982; 15:141-154.

[21] Rubin CT, Lanyon LE. Regulation of bone mass by mechanical strain magnitude. Calcif Tissue Int. 1985;37:411-417.

[22] Burr DB, Martin RB, Martin PA. Lower extremity loads stimulate bone formation in the vertebral column: implications for osteoporosis. Spine. 1983:8:681-686.

[23] Woodward MI, Cunningham JL. Skeletal accelerations measured during different exercises. Proc Inst Mech Eng [H]. 1993;207:79-85.

[24] Costigan PA, Wyss UP, Deluzio KJ, Li J. Semiautomatic three-dimensional knee motion assessment system. Med Biol Eng Comput. 1992;30:343-350.

[25] Li J, Wyss UP, Costigan PA, Deluzio KJ. An integrated procedure to assess knee-joint kinematics and kinetics during gait using an optoelectric system and standardized x-rays. J Biomed Eng. 1993;15:392-400.

[26] Siu DW, Cooke TDV, Broekhoven LD, et al. A standardized technique for lower limb radiography: practice, applications, and error analysis. Invest Radiol. 1991;26:71-77.

[27] Cooke TDV, Scudamaore RA, Bryant JT, et al. A quantitative approach to radiography of the lower limb: principles and applications. J Bone Joint Surg Br. 1991;73:715-720.

[28] Perry J. Hip. In: Perry J, Hislop HJ, eds. Gait Analysis: Normal and Pathological. Alexandria, Va: American Physical Therapy Association; 1991:111-128.

[29] Apkarian J, Naumann S, Cairs B. A three-dimensional kinematic and dynamic model of the lower limb. J Biomech. 1989;22:143-155.

[30] Eng JJ, Winter DA. Kinetic analysis of the lower limbs during walking: What information can be gained from a three-dimensional model? J Biomech. 1995;28:753-758.

[31] Norkin CC, Levangie PK. Gait. In: Norkin CC, Levangie PK, eds. Joint Structure and Function. A Comprehensive Analysis. 2nd ed. Philadelphia, Pa: FA Davis Co; 1992:448-495.

[32] Perry J. The mechanics of walking. In: Perry J, Hislop HJ, eds. Principles of Lower-Extremity Bracing. Alexandria, Va: American Physical Therapy Association; 1967:7-33.

[33] Andriacchi TP, Mikosz RP. Mnsctdoskeletal dynamics, locomotion, and clinical applications. In: Mow VC, Hayes WC, eds. Basic Orthopaedic Biomechanics. New York, NY: Raven Press; 1991:51-92.

[34] Hurwitz DE, Andriacchi TP. Biomechanics of the hip and knee. In: Nordin M, Andersson GBJ, Pope MH, eds. Musculoskeletal Disorders in the Workplace: Principles and Practice. St Louis, Mo: Mosby; 1997:487-496.

[35] Winter DA, Eng JJ, Ishac MG. A review of kinetic parameters in human walking. In: Craik RL, Oatis CA, eds. Gait Analysis: Theory and Application. St Louis, Mo: Mosby; 1995:252-270.

[36] Birge SJ. Osteoporosis and hip fracture. Clin Geriatr Med. 1993;9: 69-86.

RN Kirkwood, PhD, PT, is Physical Therapist and Clinical Coordinator at Clinical Mechanics Group, Queen's University, Kingston, Ontario, Canada.

EG Culham, PhD, PT, is Associate Professor, School of Rehabilitation Therapy, Faculty of Health Sciences, Queen's University, Kingston, Ontario, Canada K7L 3N6 ([e-mail]). Address all correspondence to Dr Culham.

P Costigan, PhD, is Assistant Professor, School of Physical and Health Education, Queen's University.

This research was conducted as part of Dr Kirkwood's doctoral degree requirements.

Financial support for this study was provided by the Medical Research Council of Canada and by CAPES (Coordenacao de Aperfeicoamento de Pessoal de Nivel Superior), Brazilia, Brazil.

Portions of this study were presented orally at the Canadian Physiotherapy Association Congress; June 26, 1998; St John's, Newtoundland, Canada.

This study was approved by the Human Research Ethics Board, Queen's University.

This article was submitted April 10, 1998, and was accepted November 16, 1998.