An aim of weight loss in obese persons is to reduce body fat and thus risk factors for metabolic syndrome (24). After weight loss, a significant amount of total daily energy expenditure is essential to maintain body weight in weight-reduced individuals (23). However, after weight loss, the energy expenditure (kJ·min−1) during walking can decrease more than one would expect from changes in body mass (15,26,31) because of the decrease in net (gross − resting) metabolic cost of walking (net CW; J·m−1) (13,20). By analyzing walking, a convenient form of daily physical activity recommended for weight management, we may gain insights into the biomechanical factors associated with the net metabolic cost of walking in obese and weight-reduced individuals.
The net metabolic cost of walking in normal-weight adults is determined by mechanical tasks such as generating force to support body weight, performing external work (W ext) to lift and accelerate the center of mass (COM), swinging the limbs, and maintaining stability (8,10,12,16,17). The muscle force generation of the lower extremity required to support body weight and the muscle work required to lift and accelerate the COM make body mass the primary determinant of net CW (J·m−1). Consequently, net CW (J·m−1) is greater in obese than that in normal-weight subjects (32). However, net CW normalized by body mass (J·kg−1·m−1) is still greater in obese than that in normal-weight subjects (4,32). Therefore, other factors probably contribute to the high metabolic cost in obese subjects (32). Previous studies have examined the external mechanical work (W ext) in obese subjects (6,27,32), which is the primary determinant of net CW (10,17). The results of these studies have shown that W ext (J·m−1) was greater in obese than that in normal-weight subjects, but the differences disappeared when W ext was normalized by body mass (J·kg−1·m−1). This suggests that body mass is the primary determinant of W ext and that the inverted pendulum mechanism of energy exchange (allowing recovery of mechanical energy from both gravitational potential and kinetic energy fluctuations) is not impaired in obese subjects (6,27,32). However, although W ext (J·kg−1·m−1) is similar between obese and normal-weight subjects, it explains about one half of the variance in net CW (J·kg−1·m−1) in adults (17). Thus, W ext (J·kg−1·m−1) and body mass may in part explain the variance in net CW (J·m−1) in obese subjects. Consequently, a decrease in body mass could be the main determinant of the decrease in net CW (J·m−1) but without change in W ext (J·kg−1·m−1) because the latter factor is not altered by obesity (6,27,32).
Other biomechanical factors that could increase the metabolic cost of walking include greater fluctuations in mediolateral (M-L) kinetic energy of the COM and greater lateral leg swing induced by the greater step width of obese subjects (27,32,38). Obese subjects are thought to walk with a wider step width to reduce the friction between the legs and/or to increase stability (38). The greater fluctuations in M-L kinetic energy of the COM (ΔEkl; J·kg−1·m−1) could require higher muscle activation and cocontraction of antagonist muscles for stabilizing the COM during M-L motions, especially during the single limb support phase. This hypothesis could explain why obese subjects typically reduce the single support duration (5,27). Indeed, decreased single support duration could reduce the energy cost of muscle force generation required to stabilize the COM and to support body weight. Furthermore, obese adults walk with wider lateral leg swing (38). Walking with an enforced wide lateral leg swing has been suggested to increase net CW in normal-weight adults by up to 30% (4). Consequently, one could expect after weight loss that weight-reduced individuals decrease step width and thus lateral leg swing and fluctuations in M-L kinetic energy. Therefore, we hypothesized that these parameters partly explain the reduction in net CW (J·m−1) observed in weight-reduced individuals.
In addition, gender differences have been observed in metabolic cost of walking (net and gross) and in its decrease after weight loss in both obese adolescents and adults (4,26). Browning et al. (4) have investigated body mass distribution in the leg region to explain the 10% greater normalized net CW in obese women, but this parameter did not explain the gender difference in net CW. However, the main morphological gender differences lie in body fat distribution, mainly located in the abdomen and gluteal-femoral regions (18). Indeed, an android type of obesity (frequently observed in men) accounts for the accumulation of adipose tissue in the abdomen, whereas a gynoid type of obesity (frequently observed in women) accounts for the accumulation of adipose tissue in the gluteal-femoral region. To our knowledge, the effect of this difference in body fat distribution on CW has not been studied. We hypothesized that net CW (J·m−1) could be partly related to fat mass in the gynoid region and that the reduction in net CW (J·m−1) with weight loss could be partly due to the decrease in fat mass in the gynoid region.
The first aim of this study was to investigate the association between biomechanical parameters of the walking gait and net CW (J·m−1). We hypothesized that the major part of the variance in net CW (J·m−1) could be explained by body mass, W ext (J·kg−1·m−1), single support duration, ΔEkl (J·kg−1·m−1), lateral leg swing (m), percent body fat, and percentage of gynoid fat. The second and primary aim was to determine whether modifications in these biomechanical parameters associated with weight loss were responsible for the greater than expected (from the change in body mass) decrease in net CW (J·m−1) in obese adolescent boys and girls at a preset walking speed. We hypothesized that this decrease (%) in net CW (J·m−1) associated with weight loss could be explained by the changes (%) in body mass, single support duration, ΔEkl (J·kg−1·m−1), lateral leg swing (m), percent body fat, and percentage of gynoid fat. Because W ext (when normalized by body mass) is not influenced by obesity, it should not be modified with weight loss and thus should not explain the decrease in net CW (J·m−1).
The present study included 16 obese adolescents (7 boys and 9 girls) who were involved in an obesity management program in the children medical center of Romagnat (CMI, Centre Médical Infantile), France. None of them was regularly performing any sporting activity or receiving any medication that could interfere with their walking pattern or influence their energetic metabolism. The main inclusion criteria were age between 12 and 16 yr and body mass index (BMI, kg·m−2) above age- and gender-specific cutoff points for obesity as defined by Cole et al. (9).
Subjects were housed at the medical center (except during the weekends that were spent at home) where they underwent a 12-wk voluntary weight reduction program, including nutritional education, caloric restriction, and physical activities, which consisted of 40-min sessions of aerobic fitness, strength training, and supervised free practice per week. Diet composition was formulated according to the French-recommended dietary allowances (28), and on average, subjects lost 1 kg per week before a stabilization phase that lasted about 2 wk before leaving the center. Data of subjects who had lost less than 3% of body mass have been excluded from further analyses. The physical characteristics of the subjects before and after weight loss are presented in Table 1.
For each adolescent and his or her parents, the study was explained in detail, and the written consent was obtained before the beginning of the study. This study was approved by the regional ethics committee and was performed in accordance with the Declaration of Helsinki II.
Subjects were tested twice in the same conditions: the first test was done before weight loss on the first or second day of the obesity management program, and the second was done during the last week of the stabilization phase. Body composition was assessed on the day of each test or on the day before.
For each subject, the standing rate of oxygen consumption (V˙O2) was first measured over 10 min. Then, all subjects performed five 4-min tests, walking along an athletic track lane (with two straight lines of 25 m), at different walking speeds (0.75, 1, 1.25, and 1.5 m·s−1 and at preferred walking speed) in a randomized order, separated by 5 min of rest. The slope of the track was tested every 1 m and ranged from −0.5% to +0.5 %. The walking speed was controlled using markers set out every 5 m along the track, and the subjects were instructed to walk past the markers at a pace imposed by a metronome tone. An experimenter walked alongside each subject to help him or her match the required speed. Mechanical and metabolic parameters of walking were measured with two portable devices carried around the chest by the subjects. Given the large number of parameters measured, only the parameters recorded at 1.25 m·s−1 (which is close to the mean preferred walking speed) were retained for analysis.
Assessment of Body Composition
Regional (arms, legs, trunk, gynoid, and android) and total body fat and lean body mass (LBM; the mass of nonbone lean tissue) were measured by dual-energy x-ray absorptiometry (QDR 4005; Hologic Inc., Bedford, MA). The arms, the legs, and the trunk regions were delineated with the use of specific anatomical landmarks as previously described by Berends et al. (2). The android region was defined inferiorly at the pelvis cut line and superiorly above the pelvis cut line by 20% of the distance between the pelvis and the neck cut (2). The gynoid region was defined superiorly at the pelvis cut line and inferiorly by a transverse cut line at the level of the lesser trochanters (Fig. 1). Percentages of total and regional body fat were calculated by dividing total and regional body fat mass by total and regional body mass, respectively. The type of obesity (android or gynoid) was estimated from the ratio of android fat (%) to gynoid fat (%).
For all subjects, stature was measured to the nearest 0.5 cm using a standardized wall-mounted height board, and BMI was calculated as body mass divided by height squared.
Assessment of Metabolic Parameters
V˙O2 (mL·min−1) and the rate carbon dioxide production (V˙CO2; mL·min−1) were measured using a breath-by-breath portable gas analyzer (K4b2, COSMED s.r.l., Italy) that weighed less than a kilogram and recorded and stored the data for the entire session for each subject. The K4b2 unit, previously validated by Duffield et al. (14), was calibrated with standard gases before each session. Average V˙O2 and V˙CO2 were calculated over 30 s taken during the last minute of the trial where V˙O2 and V˙CO2 were stable within ±10%. Gross metabolic rate (W) for each 4-min test and standing metabolic rate (W) were assessed from the steady-state V˙O2 and V˙CO2 using Brockway's (3) standard equation. Standing metabolic rate (W) was divided by body mass and LBM to obtain normalized standing metabolic rate in watts per kilogram and watts per kilogram of LBM, respectively. Gross metabolic rate (W) was divided by walking speed (m·s−1) to obtain gross CW (J·m−1) and finally by body mass to obtain normalized gross CW (J·kg−1·m−1). Net metabolic rate (W) was calculated by subtracting standing metabolic rate (W) from gross metabolic rate (W). Then, net metabolic rate (W) was divided by walking speed (m·s−1) to obtain net CW (J·m−1) and finally by body mass to obtain normalized net CW (J·kg−1·m−1).
Assessment of Mechanical Parameters
Mechanical parameters of walking were calculated from the three-dimensional accelerations of two inertial sensors equipped with a triaxial (three orthogonal axes) accelerometer and gyroscope (MTx, Xsens, Enschede, The Netherlands). As described by Pfau et al. (33), these inertial sensors have been validated for determining the position of an object from the sensor three-dimensional linear accelerations during cyclical movement.
Thus, in accordance with the validation study by Meichtry et al. (29) and as previously described by Peyrot et al. (32), an inertial/gyroscope sensor (∼0.03 kg) was taped and secured directly to the skin on the lower part of the back, facing the L3 vertebra region (close to the COM) using an adhesive strap, and was used to measure the COM accelerations. It was assumed that changes in the relative positions of the sensor and COM over time during walking may be neglected. Then, a second sensor was also taped and secured on the instep of subjects' right foot to measure three-dimensional foot accelerations and in turn to assess lateral leg swing from foot displacements during the swing phase of the right leg. Sabatini et al. (36) have shown that an inertial/gyroscope sensor placed on the instep of the foot allows to accurately reconstruct the sagittal trajectory of this anatomical point.
The sensors were connected to a lightweight (∼0.3 kg) data logger carried by the subjects at the middle of the thoracic spine. Data were sampled at 100 Hz and transmitted to a computer by telemetry over the 4 min of the trial.
As described in detail by Pfau et al. (33), the orientation algorithm of the inertial sensors provided orientation data in the earth reference system (horizontal and magnetic north) in the form of Euler angles (roll, pitch, and heading). Euler angles represent rotations of the sensor system into the earth reference system, with the magnetic north corresponding to the anteroposterior axis in our study. Thus, rotation matrices were used to reposition three-dimensional accelerations of the two sensors in the earth reference system. Data were low-pass filtered at 30 Hz (fourth-order, zero-lag, low-pass Butterworth).
Heel strike and toe-off were determined from forward and vertical right foot accelerations peaks, as described in detail by Jasiewicz et al. (22). Thus, stride duration (delimited by two consecutive heel strikes) and stance duration (heel strike to consecutive toe-off) were computed as well as single support duration of the contralateral limb (i.e., right toe-off to right consecutive heel strike). Then, stance and single support durations were expressed relatively to stride duration (%). Stride frequency (Hz) was computed as the inverse of stride duration.
COM energy fluctuations and external mechanical work.
A mean stride was obtained by averaging the three-dimensional COM accelerations of five time-normalized consecutive strides taken during the 30-s period when metabolic data were collected. Then, three-dimensional accelerations of the mean stride were integrated twice to obtain vertical, forward, and M-L COM velocities and positions, as proposed by Cavagna (7).
The total instantaneous kinetic (Ek; J), M-L kinetic (Ekl; J), and potential (Ep; J) energies of the COM were calculated as follows:
where m is the body mass (kg); V is the resultant COM velocity (m·s−1) determined from its vertical, forward, and M-L components; Vl is the M-L COM velocity (m·s−1); g is the gravitational constant; and h is the vertical position of the COM. The total mechanical energy of the COM (E tot; J) was computed as the sum of the Ek and Ep curves over the mean stride. W ext (J) was calculated as the sum of the positive increments in E tot and divided by stride length and then by body mass to be expressed in joules per meter and in joules per kilogram per meter, respectively. The inverted pendulum recovery of mechanical energy of the COM was calculated according to Schepens et al. (37) as follows:
where ΔEk (J) and ΔEp (J) are the fluctuations in kinetic and potential energy of the COM, calculated as the sum of the positive increments in Ek and Ep, respectively.
Fluctuations in M-L kinetic energy (ΔEkl; J) were also calculated as the sum of the positive increments in Ekl. Finally, ΔEk, ΔEkl, and ΔEp were divided by stride length and then by body mass to be expressed in joules per meter and in joules per kilogram per meter, respectively.
Lateral leg swing.
Lateral leg swing of the right leg was assessed from foot displacements in the transverse plane. M-L mean foot accelerations were obtained by averaging five time-normalized consecutive M-L foot accelerations during the swing periods. Then, these M-L mean accelerations were integrated twice to obtain M-L foot positions. Lateral leg swing (m) was defined as the M-L amplitude, from medial to lateral, of the foot position during the right leg swing phase.
Normal distribution of the data was checked by the Shapiro-Wilk normality test. Variance homogeneity between samples was tested by the Snedecor F-test. A two-way (period and gender) ANOVA with repeated measures was used to determine the effects of the period (before and after the weight reduction program), gender and their interaction (period × gender) on body composition, and metabolic and mechanical parameters. If an interaction between weight loss period and gender was significant, unpaired (boys vs girls before and after weight loss) and paired (before vs after weight loss for boys and girls) t-tests were performed.
A first multiple linear regression analysis was performed on the entire sample (pooling the data of pre- and postweight loss conditions; n = 32) to determine the fraction of the variance in net CW(J·m−1) attributable to body mass, W ext (J·kg−1·m−1), single support duration, ΔEkl (J·kg−1·m−1), lateral leg swing, percent body fat, and percentage of gynoid fat. Beforehand, a correlation matrix had been performed between independent variables entered together into the multiple regression analysis, showing significant correlations between body mass and percent body fat and between percent body fat and percentage of gynoid fat. To examine the effects of these variables independently, percent body fat was normalized by body mass, and percentage of gynoid fat was normalized by percent body fat. Because regression analysis involved linear regressions, body mass has been raised to the 0.67 power (32,34,40).
A second multiple regression analysis (n = 16) was performed to determine the fraction of the variance in the changes in net CW with weight loss attributable to changes in body mass, single support duration, ΔEkl (J·kg−1·m−1), lateral leg swing, percent body fat, percentage of gynoid fat, and ΔEp (J·kg−1·m−1). The latter was entered in the multiple regression analysis because it was the only mechanical parameter that unexpectedly changed with weight loss.
Then, a backward elimination criterion was used to eliminate variables that did not explain a significant amount of the variance in the dependent variable. Lastly, the adjusted r 2, which takes into account the number of independent variables included in the multiple regression model, was retained to account for the fraction of the variance in the dependent variable (net CW) attributable to changes in the independent variables (biomechanical parameters). The criterion for statistical significance was set at P < 0.05.
The weight loss intervention program was effective for both boys and girls (Table 1). Body composition was not significantly different between boys and girls. LBM tended to be higher in boys than that in girls (P = 0.052) but did not change with weight loss in both groups.
Fat percentage in the different regions (trunk, android, and gynoid) decreased significantly in both boys and girls after weight loss (Table 2), except for fat percentage in the legs in girls (t-test, P = 0.25). There was no significant difference in fat percentage in the android and gynoid region between boys and girls. Android-to-gynoid fat ratio did not change significantly with weight loss and was not significantly different between boys and girls (Table 2).
Metabolic Cost of Walking
Net CW (J·m−1) and net CW normalized per kilogram of body mass decreased significantly after weight loss in both boys and girls (Table 3). There was no significant difference between boys and girls in net CW (J·m−1 or J·kg−1·m−1) and in its changes associated with weight loss.
The first multiple linear regression analysis used to put forward the determinants of net CW (J·m−1) contained five variables all positively correlated: body mass, W ext (J·kg−1·m−1), single support duration, percent body fat, and percentage of gynoid fat. This five-variable model explained over 77% of the variance in net CW (J·m−1) (r = 0.90, F = 22.1, P < 0.001, n = 32). Lateral leg swing and ΔEkl (J·kg−1·m−1) did not explain a significant part of the variance in net CW (J·m−1).
The second multiple linear regression analysis used to put forward the determinants of net CW changes associated with weight loss contained six variables positively correlated: the changes in body mass, ΔEkl (J·kg−1·m−1), ΔEp (J·kg−1·m−1), single support duration, percent body fat, and percentage of gynoid fat. This six-variable model explained 65% of the changes in net CW with weight loss (r = 0.89, F = 5.6, P < 0.012, n = 16).
There was no significant difference in spatiotemporal parameters between boys and girls (Table 4). Stance duration (%) and single support duration (%) did not change significantly with weight loss. Stride length increased significantly in boys and girls after weight loss and was positively correlated to subjects' stature (r = 0.58, P < 0.01). There was no significant difference in lateral leg swing (m) between boys and girls. Lateral leg swing decreased significantly after weight loss, yet this decrease was not related to changes in mass or in body composition of the lower limbs after weight loss.
COM energy fluctuations and external mechanical work.
There was no significant difference between boys and girls for all kinetic parameters of walking. When expressed in joules per meter, ΔEk, ΔEkl, and ΔEp were significantly reduced after weight loss (−8.5% ± 9.0%, −22.9% ± 17.0%, and −11.4% ± 9.9%, respectively; Table 5). When expressed per kilogram of body mass, ΔEk did not decrease significantly after weight loss, whereas ΔEkl and ΔEp did (J·kg−1·m−1; −18.1% ± 17.8% and −5.8% ± 11.0%, respectively; Table 5). W ext (in both J·m−1 and J·kg−1·m−1) and recovery of mechanical energy did not change significantly after weight loss.
The results of this study indicate that in obese subjects, net CW (J·m−1) is related to biomechanical parameters of walking. More importantly, after weight loss, the greater than expected (from the change in body mass) decrease in net CW (J·m−1) is associated with changes in biomechanical parameters of walking. However, contrary to what we hypothesized, net CW (J·m−1) was not related to lateral leg swing, and the reduction in net CW (J·m−1) associated with weight loss was not related to the decrease in lateral leg swing. On the other hand, 77% of the variance in net CW (J·m−1) was explained by the variance in body mass, W ext (J·kg−1·m−1), single support duration, percent body fat, and percentage of gynoid fat. Moreover, 65% of the changes in net CW (J·m−1) with weight loss were explained by changes in body mass, ΔEkl (J·kg−1·m−1), ΔEp (J·kg−1·m−1), single support duration, percent body fat, and percentage of gynoid fat.
The greater than expected (from the change in body mass) decrease in net CW (J·m−1) in obese adolescents after weight loss was consistent with the results of previous studies in obese adults (13,20). In the present study, net CW (J·m−1) decreased by 13.3% ± 15.4%, whereas body mass decreased by 6.0% ± 2.2%, which induced a 7.8% ± 16.0% decrease in normalized net CW (J·kg−1·m−1). Our metabolic and mechanical values are consistent with the values of prior studies. Interpolated values of gross metabolic rate (kJ·min−1) at 1.1 and 1.4 m·s−1 in the present study were similar to those reported by Lazzer et al. (25) in obese adolescents with similar body mass and composition. Moreover, values of mechanical parameters are in agreement with those recently obtained by Malatesta et al. (27) in obese adults. For example, they reported a mean value of W ext of 39.5 J·m−1 at the mean speed of 1.18 m·s−1, which is close to the 38.9 J·m−1 measured in the present study before the weight loss program at 1.25 m·s−1. Lateral leg swing could not be compared with other studies because no such data in obese subjects are available to our knowledge.
Contrary to our hypotheses, net CW (J·m−1) was not related to lateral leg swing, and the reduction in net CW (J·m−1) associated with weight loss was not related to the decrease in lateral leg swing (∼11%). Gottschall and Kram (16) have shown in normal-weight subjects that the metabolic cost of leg swing comprises only between 10% and 15% of the net CW at 1.25 m·s−1. It is therefore possible that a wider leg swing in obese individuals could induce no change or only a moderate increase in net CW, as shown for heavier legs (4). As a consequence, the lateral leg swing decrease seems not to be involved in the reduction in net CW (J·m−1) associated with weight loss.
As hypothesized, net CW (J·m−1) was related to body mass, and their respective decreases with weight loss were also related. The effect of body mass on net CW (J·m−1) results from leg muscle work required to raise and accelerate this mass. Moreover, normalized W ext (J·kg−1·m−1), which accounts for the mechanical work required to move the COM independently of body mass, also explained part of the variance in net CW (J·m−1) in obese adolescents. This result is consistent with the findings of Donelan et al. (10) in normal-weight adults. Consequently, in obese subjects, the higher the body mass and the mechanical work required to move the COM per unit body mass, the higher the net CW (J·m−1). As expected, W ext (J·kg−1·m−1) did not change with weight loss and was not involved in the decrease in net CW (J·m−1). Moreover, single support duration, during which one single leg supports body weight, was related to net CW(J·m−1). Changes in single support duration were also related to changes in net CW (J·m−1) with weight loss. It has been reported that obese subjects walk with shorter single support durations and with greater sagittal-plane hip, knee, and ankle net muscle moments during this phase (5). It is therefore possible that obese subjects decrease the duration of single limb support to decrease the duration of the higher muscle activation required to support body weight. In addition, obesity could increase the cost of supporting body weight because of the reduced relative strength in obese compared with normal-weight adults (19). This reduced strength could require a relative increase in muscle activation, and in the recruitment of fast glycolytic fibers to support body weight, these fibers being less economical than slow oxidative ones (21). Thus, in weight-reduced individuals, changes in single support duration and hence in muscle activation could partly explain the changes in net CW (J·m−1) with weight loss.
Although normalized W ext (J·kg−1·m−1) did not change with weight loss, normalized ΔEp (J·kg−1·m−1) and ΔEkl (J·kg−1·m−1) decreased with weight loss. Even if it did not significantly change, ΔEk (J·kg−1·m−1) tended to decrease with weight loss. Walking fundamentally involves an inverted pendulum-like mechanism that conserves mechanical energy. Vertical motions of the COM and hence ΔEp (J·kg−1·m−1) allow pendulum-like exchange between potential and kinetic energy (8). Thus, it is possible that weight-reduced individuals decreased vertical movement and hence ΔEp (J·kg−1·m−1) partly because of the decrease in ΔEk (J·kg−1·m−1) with weight loss, specially in the M-L direction. The decrease in ΔEkl (J·kg−1·m−1) could be related to a decrease in step width associated with an increase in stability and/or a decrease in thigh circumference with weight loss. The fact that the decrease in ΔEp (J·kg−1·m−1) could be related to the decrease in ΔEk (J·kg−1·m−1) is supported by the similar ratio of ΔEk /ΔEp before and after weight loss (1.12 vs 1.15, respectively; P = 0.28). We can note that our values of ΔEk /ΔEp are in agreement with those recently obtained at similar speed by Malatesta et al. (27) in obese adults. However, our results have shown a greater stride length after weight loss, which could be due to the normal adolescent development during the 12-wk period. It is therefore possible that the change in lower extremity kinematics after weight loss is also partly responsible for the change in ΔEp (J·kg−1·m−1) (10).
The decreases in ΔEp (J·kg−1·m−1) and ΔEkl (J·kg−1·m−1) were also both involved in the decrease in net CW (J·m−1) associated with weight loss. Indeed, although vertical motion allows inverted pendulum energy exchange and therefore reduces the mechanical work required to accelerate the COM, some evidence suggests that the metabolic cost of raising the COM is significant (30). Neptune et al. (30) have shown that significant muscle work is needed to raise the COM by extending the knee and hip and increasing the COM potential energy. Consequently, the decrease in ΔEkl (J·kg−1·m−1) and therefore in ΔEp (J·kg−1·m−1) could have induced a lesser muscle force generation to raise the COM and hence explained part of the decrease in net CW (J·m−1) associated with weight loss. In addition, the decrease in ΔEkl (J·kg−1·m−1) could have induced a smaller level of muscle activation and cocontraction of antagonist muscles for stabilizing the COM during M-L motions, especially during the single limb support phase.
Regarding body fat distribution, obese adolescent boys and girls presented similar mean values of percentage of fat in android and gynoid regions, which could be due to the pubertal status of the subjects (1,18). Indeed, He et al. (18) have shown that gender differences in gynoid fat distribution were evident in prepuberty and late puberty but absent in early puberty, mostly because of the variations in body fat distribution in boys during puberty. Although net CW (J·m−1) was related to the percentage of gynoid fat, the absence of gender effect on CW (J·m−1) could be due to the differences in pubertal status and hence in obesity status (android or gynoid) within populations of boys and girls. This idea is supported by the similar android-to-gynoid fat ratio observed in our study between boys and girls. However, we can assume that the greater net metabolic rate (W·kg−1) in obese adult women reported by Browning et al. (4) could be due to the gynoid obesity usually reported in women (18). In addition, our results showed that the decrease in net CW (J·m−1) was partly related to the decrease in percentage of gynoid fat. However, the physiological and/or biomechanical effects of fat tissue in the gynoid region, which could induce gender differences in net CW (J·m−1), are still not fully understood. It is possible that fat tissue in the gynoid region induced changes in other biomechanical parameters of walking.
Recently, it has also been suggested that an increase in efficiency of muscle mechanical work with weight loss could explain part of the decrease in energy expended in physical activity (35). Thus, it is possible in our study that the remaining variance of the relation between the changes in net CW (J·m−1) and the changes in the biomechanical parameters might be explained by an increase in efficiency of muscle mechanical work with weight loss. Measuring total mechanical work and metabolic cost of walking simultaneously would provide valuable insights into the change in efficiency of muscle mechanical work during walking associated with weight loss. Total mechanical work would be calculated as the sum of W ext (which includes the mechanical work done by one leg against the other during the double contact phase) plus the work required to accelerate the body segments relative to the COM (11,39).
Some methodological limitations should be addressed in this study, such as the use of inertial/gyroscope sensors. Indeed, although accelerometry is sufficiently accurate for assessing mechanical parameters of the walking gait (29,32,36), there are possible sources of approximations such as sensor movements induced by skin movements or variations in the positions of the sensor relatively to the COM. In addition, the use of inertial/gyroscope sensors did not allow us to quantify the mechanical work done by one leg against the other during the double contact phase. Consequently, it is not possible to determine whether weight loss induced any change in the amount of work done by the individual limbs. Moreover, normal adolescent development may have influenced the outcomes of this study. Indeed, adolescents, especially boys, grew taller during the 12-wk period, which could have interfered with the weight loss program itself. We can reasonably assume that without this growth effect (obviously inevitable), the weight loss program could have induced a greater decrease in body mass. Moreover, in addition to weight loss, the 1-cm (0.5%) increase in the adolescents' height may have induced changes in their walking mechanics. However, we can note that although adolescents grew taller during the 12-wk period, this has not induced any change in subjects' preferred walking speed (before and after weight loss, 1.25 ± 0.14 vs 1.24 ± 0.15; P = 0.77).
It is concluded that in obese adolescents, the reduction in net metabolic cost after weight loss is associated with changes in the biomechanical parameters of walking. The smaller lateral leg swing after weight loss does not seem to explain part of the decrease in net metabolic cost. The main determinant of the decrease in net metabolic cost (J·m−1) is body mass, which likely reduces the leg muscle work required to raise and accelerate the COM as well as to support body weight. The decrease in body mass after weight may be associated with a lesser leg muscle work required to raise the COM because of smaller vertical motions. This decrease in vertical motion seems to be related to the decrease in mediolateral kinetic energy fluctuations. Indeed, as vertical motions allow pendulum-like exchange between potential and kinetic energy, weight-reduced individuals could reduce the potential energy available (hence vertical motions) because of the decrease in mediolateral kinetic energy fluctuations. Moreover, the reduction in fat mass in the gynoid region, independently of the decrease in total body fat mass, seems related to the decrease in net metabolic cost of walking. Future studies are needed to confirm the relative importance of each of the mechanical parameters put forward here on the reduction in metabolic cost of walking in weight-reduced individuals.
This research program was supported by the French Auvergne and Rhône-Alpes Regions, thanks to the European Regional Development Funds (ERDF).
Disclosure statement: No fund was received for this study from the National Institutes of Health, the Wellcome Trust, the Howard Hughes Medical Institute, or others.
Conflict of interest: The authors declare that they have no conflict of interest.
The authors thank the subjects for their commitment during this study and the staff of the "Centre Médical Infantile" (CMI) of Romagnat for their collaboration on this study.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Aucouturier J, Meyer M, Thivel D, Taillardat M, Duche P. Effect of android to gynoid fat ratio on insulin resistance in obese youth. Arch Pediatr Adolesc Med
2. Berends AL, Zillikens MC, de Groot CJ, et al. Body composition by dual-energy x-ray absorptiometry in women with previous pre-eclampsia or small-for-gestational-age offspring. BJOG
3. Brockway JM. Derivation of formulae used to calculate energy expenditure in man. Hum Nutr Clin Nutr
4. Browning RC, Baker EA, Herron JA, Kram R. Effects of obesity and sex on the energetic cost and preferred speed of walking. J Appl Physiol
5. Browning RC, Kram R. Effects of obesity on the biomechanics of walking at different speeds. Med Sci Sports Exerc
6. Browning RC, McGowan CP, Kram R. Obesity does not increase external mechanical work per kilogram body mass during walking. J Biomech
7. Cavagna GA. Force platforms as ergometers. J Appl Physiol
8. Cavagna GA, Thys H, Zamboni A. The sources of external work in level walking and running. J Physiol
9. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ
10. Donelan JM, Kram R, Kuo AD. Mechanical work for step-to-step transitions is a major determinant of the metabolic cost of human walking. J Exp Biol
. 2002;205(pt 23):3717-27.
11. Donelan JM, Kram R, Kuo AD. Simultaneous positive and negative external mechanical work in human walking. J Biomech
12. Donelan JM, Shipman DW, Kram R, Kuo AD. Mechanical and metabolic requirements for active lateral stabilization in human walking. J Biomech
13. Doucet E, Imbeault P, St-Pierre S, et al. Greater than predicted decrease in energy expenditure during exercise after body weight loss in obese men. Clin Sci (Lond)
14. Duffield R, Dawson B, Pinnington HC, Wong P. Accuracy and reliability of a Cosmed K4b2
portable gas analysis system. J Sci Med Sport
15. Foster GD, Wadden TA, Kendrick ZV, Letizia KA, Lander DP, Conill AM. The energy cost of walking before and after significant weight loss. Med Sci Sports Exerc
16. Gottschall JS, Kram R. Energy cost and muscular activity required for leg swing during walking. J Appl Physiol
17. Grabowski A, Farley CT, Kram R. Independent metabolic costs of supporting body weight and accelerating body mass during walking. J Appl Physiol
18. He Q, Horlick M, Thornton J, et al. Sex-specific fat distribution is not linear across pubertal groups in a multiethnic study. Obes Res
19. Hulens M, Vansant G, Lysens R, Claessens AL, Muls E, Brumagne S. Study of differences in peripheral muscle strength of lean versus obese women: an allometric approach. Int J Obes Relat Metab Disord
20. Hunter GR, McCarthy JP, Bryan DR, Zuckerman PA, Bamman MM, Byrne NM. Increased strength and decreased flexibility are related to reduced oxygen cost of walking. Eur J Appl Physiol
21. Hunter GR, Newcomer BR, Larson-Meyer DE, Bamman MM, Weinsier RL. Muscle metabolic economy is inversely related to exercise intensity and type II myofiber distribution. Muscle Nerve
22. Jasiewicz JM, Allum JH, Middleton JW, et al. Gait event detection using linear accelerometers or angular velocity transducers in able-bodied and spinal-cord injured individuals. Gait Posture
23. Kinney JM. Influence of altered body weight on energy expenditure. Nutr Rev
24. Kukkonen-Harjula KT, Borg PT, Nenonen AM, Fogelholm MG. Effects of a weight maintenance program with or without exercise on the metabolic syndrome: a randomized trial in obese men. Prev Med
25. Lazzer S, Boirie Y, Bitar A, et al. Assessment of energy expenditure associated with physical activities in free-living obese and nonobese adolescents. Am J Clin Nutr
26. Lazzer S, Boirie Y, Montaurier C, Vernet J, Meyer M, Vermorel M. A weight reduction program preserves fat-free mass but not metabolic rate in obese adolescents. Obes Res
27. Malatesta D, Vismara L, Menegoni F, Galli M, Romei M, Capodaglio P. Mechanical external work and recovery at preferred walking speed in obese subjects. Med Sci Sports Exerc
28. Martin A. The "apports nutritionnels conseilles (ANC)" for the French population. Reprod Nutr Dev
29. Meichtry A, Romkes J, Gobelet C, Brunner R, Muller R. Criterion validity of 3D trunk accelerations to assess external work and power in able-bodied gait. Gait Posture
30. Neptune RR, Zajac FE, Kautz SA. Muscle mechanical work requirements during normal walking: the energetic cost of raising the body's center-of-mass is significant. J Biomech
31. Ohrstrom M, Hedenbro J, Ekelund M. Energy expenditure during treadmill walking before and after vertical banded gastroplasty: a one-year follow-up study in 11 obese women. Eur J Surg
32. Peyrot N, Thivel D, Isacco L, Morin JB, Duche P, Belli A. Do mechanical gait parameters explain the higher metabolic cost of walking in obese adolescents? J Appl Physiol
33. Pfau T, Witte TH, Wilson AM. A method for deriving displacement data during cyclical movement using an inertial sensor. J Exp Biol
. 2005;208(Pt 13):2503-14.
34. Rogers DM, Olson BL, Wilmore JH. Scaling for the VO2
-to-body size relationship among children and adults. J Appl Physiol
35. Rosenbaum M, Vandenborne K, Goldsmith R, et al. Effects of experimental weight perturbation on skeletal muscle work efficiency in human subjects. Am J Physiol Regul Integr Comp Physiol
36. Sabatini AM, Martelloni C, Scapellato S, Cavallo F. Assessment of walking features from foot inertial sensing. IEEE Trans Biomed Eng
37. Schepens B, Bastien GJ, Heglund NC, Willems PA. Mechanical work and muscular efficiency in walking children. J Exp Biol
. 2004;207(pt 4):587-96.
38. Spyropoulos P, Pisciotta JC, Pavlou KN, Cairns MA, Simon SR. Biomechanical gait analysis in obese men. Arch Phys Med Rehabil
39. Willems PA, Cavagna GA, Heglund NC. External, internal and total work in human locomotion. J Exp Biol
. 1995;198(pt 2):379-93.
40. Zakeri I, Puyau MR, Adolph AL, Vohra FA, Butte NF. Normalization of energy expenditure data for differences in body mass or composition in children and adolescents. J Nutr