Body weight is a strong predictor of postural stability

Gait & Posture 26 (2007) 32–38 www.elsevier.com/locate/gaitpost

Body weight is a strong predictor of postural stability Olivier Hue a, Martin Simoneau a, Julie Marcotte a, Fe´lix Berrigan a, Jean Dore´ a, Picard Marceau b,c, Simon Marceau b,c, Angelo Tremblay a, Normand Teasdale a,* a

Faculty of Medicine, Department of Social and Preventive Medicine, Division of Kinesiology, Laval University, PEPS, Que´bec G1K 7P4, Canada b Faculty of Medicine, Department of Surgery, Laval University, Que´bec, Canada c Department of General Surgery, Laval Hospital, Que´bec, Canada Received 3 March 2006; received in revised form 3 July 2006; accepted 17 July 2006

Abstract Proper balance control is a key aspect of acitivities of daily living. The aim of this study was to determine the contribution of body weight to predict balance stability. The balance stability of 59 male subjects with BMI ranging from 17.4 to 63.8 kg/m2 was assessed using a force platform. The subjects were tested with and without vision. A stepwise multiple regression analysis was used to determine the independent effect of body weight, age, body height and foot length on balance stability (i.e., mean speed of the center of foot pressure). With vision, the stepwise multiple regression revealed that body weight accounted for 52% of the variance of balance stability. The addition of age contributed a further 3% to explain balance control. Without vision, body weight accounted for 54% of the variance and the addition of age and body height added a further 8% and 1% to explain the total variance, respectively. The final model explained 63% of the variance. A decrease in balance stability is strongly correlated to an increase in body weight. This suggests that body weight may be an important risk factor for falling. Future studies should examine more closely the combined effect of aging and obesity on falling and injuries and the impact of obesity on the diverse range of activities of daily living. # 2006 Elsevier B.V. All rights reserved. Keywords: Balance control; Body weight; Age; Center of pressure; Posturography

1. Introduction Obesity is associated with serious medical complications that impair quality of life [1–3]. It also modifies body geometry, increases the mass of the different segments [4,5], and imposes functional limitations pertaining to the biomechanics of activities of daily living that may predispose the obese to injury [6]. One of these limitations relates to balance control. Proper balance control is a critical factor in terms of fall prevention because balance impairments have been identified as important risk factors for falling [7]. The suggestion was made that, when an obese person is submitted to a small and normal forward oscillation, an abnormal distribution of body fat in the abdominal area * Corresponding author. Tel.: +1 418 656 2147; fax: +1 418 656 2441. E-mail address: [email protected] (N. Teasdale). 0966-6362/$ – see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2006.07.005

(centre of mass position relative ankle joint) yields to an increased restabilizing ankle torque needed to regain balance [8]. This suggests that, when submitted to daily postural stresses and perturbations, obese persons, particularly those with an abnormal distribution of body fat in the abdominal area, may be at higher risk of falling than lightweight individuals because they have to generate ankle torque more rapidly and with a much higher rate of torque development to recover balance. There are a number of studies with obese boys providing support to this association between body weight and balance stability. For obese boys aged 10–21, Goulding et al. [9] reported a significant negative relationship between body weight, body mass index, percentage of fat and total fat mass and a clinical balance score (Bruininks– Oseretsky). Compared to non-obese prepubertal boys, obese boys also showed greater sway areas and variability in the medial/lateral direction [10]. In addition, more obese

O. Hue et al. / Gait & Posture 26 (2007) 32–38

children suffer from traumatic accidents to anterior teeth than non-obese children suggesting that they experience more forward fall [11]. Altogether, these studies support the idea that overweight can lead to poorer balance control. With young adults, Chiari et al. [12] investigated more closely the relationship between body sway oscillations and various anthropometric parameters. Their subjects had body mass index ranging from 17.8 to 31.0 kg/m2 (mean body weight of 65.6  13.6 kg), which corresponds to an underweight/overweight range [1]. Body weight correlated with mean speed of the center of pressure (r = 0.43 for conditions with vision). Because simple correlations were used, it is not possible to determine the unique contribution of body weight independent of other anthropometric parameters likely to contribute to this relationship. For instance, body height and foot length also have been suggested to affect balance control [12,13]. More recently, Teasdale et al. [14] showed that measures of postural stability (i.e., CP speed and range in antero-posterior and lateral axes) were improved in obese and morbid obese men after losing weight (mean weight loss of 12.3 and 71.3 kg, respectively). These authors observed a strong linear relationship between the magnitude of the weight loss and the improvement in balance control thus providing support to the suggestion that weight could be an important predictor of postural stability. Therefore, the aim of the present study was to further investigate the relative contribution of body weight to balance control using a stepwise multiple regression model. In addition to body weight, body height, foot length, and age were included in the statistical model.

2. Methods 2.1. Subjects and postural stability protocol Fifty-nine male individuals were tested in this study (age range 24–61 years). Table 1 presents physical characteristics of the subjects and shows that they covered a wide range of anthropometric characteristics, particularly for Table 1 Physical characteristics of subjects Characteristics

Individuals (n = 59)

Age (years) Body height (cm) Foot length (cm)a Body weight (kg) BMI (kg/m2) Waist circumference (cm)b Hip circumference (cm) Hip/waist circumference ratio

40.5  9.5 [24–61] 175.0  6.3 [158.8–190.5] 26.8  1.2 [24.4–30.2] 107.7  35.6 [59.2–209.5] 35.2  11.7 [17.4–63.8] 115.0  28.9 [74.8–180.0] 117.1  21.2 [87.3–169.5] 1.04  0.11 [0.80–1.27]

Values are means  S.D. and [range]. a Measured between calcaneus posterior tuberosity and hallux anterior border. b Measured at mid distance between 12th rib and iliac crest.

33

body weight and BMI (59.2–209.5 kg and 17.4–63.8 kg/m2, respectively). Subjects with known neurological disorders and cognitive impairments were excluded. By definition, all morbid obese subjects had some health problems (for instance, sleep disorders, diabetes, hypertension, etc.) [1]. They all were awaiting a surgical intervention within 3 months (bariatric surgery) from testing. All participants gave their written informed consent to participate in this study, which was approved by the Laval University and the Laval Hospital Research Centre Medical Ethics Committees. Balance stability was evaluated with a force platform (Kistler model 9284). Subjects stood barefoot on the platform with their feet 10 cm apart. Before the first trial, position of the feet was traced on a sheet of paper secured on the platform. The subjects were asked to maintain a stable posture while fixating a reference point located at eye level (5 m in front of them). The arms were held alongside the body. They performed seven trials with vision and seven trials without vision (eyes closed). All trials lasted 35 s and were initiated with the eyes opened. For the no-vision condition, an auditory signal, presented 5 s before the trial, indicated the subject to close his eyes. Visual conditions were randomly presented. Subjects were able to rest midway through the experiment. An assistant helped throughout the session to ensure that procedures were adequately followed and that foot position was constant across all trials. 2.2. Data analyses Antero-posterior and medio-lateral coordinates of the center of foot pressure (CP) were determined from the ground reaction force and moments recorded at 200 Hz (12-bit A/D conversion). Prior to computing the CP displacement, the moments and force data were digitally filtered (Butterworth fourth-order, 7 Hz low-pass cut-off frequency with dual-pass to remove phase shift). The last 30 s of each trial were kept for the analyses. Mean CP speed, which corresponds to the cumulative distance over the sampling period, was used to evaluate the ability of participants to control their balance. CP speed constitutes a good index of the activity required to maintain stability [15,16] and has been considered as a sensitive and discriminant variable of postural stability [17,18]. Balance control is often described with other dependent variables. For this reason, we also extracted the following variables from the CP signals: sway area, root-mean-square of the CP position along antero-posterior (AP) and medio-lateral (ML) axes (RMS ML and RMS AP), root-mean-square of the CP velocity along ML and AP axes (RMS ML velocity and RMS AP velocity), and range of CP displacement (range ML and range AP). The sway area was computed as the elliptical area of 85% confidence of the CP displacements. RMS (both for CP displacement and velocity) are often taken as measure of stability with greater RMS values indicating less stable subjects [19]. They simply represent the standard deviation of

34

O. Hue et al. / Gait & Posture 26 (2007) 32–38

the CP displacement and velocity time series, respectively. The range of the CP displacement indicates the difference between the minimal and maximal excursions of the CP from the base of support. Finally, two parameters derived from a sway density plot approach, i.e. mean peaks and mean distances, were computed [17,20]. The peaks correspond to time instants in which the CP is relatively stable and a shorter mean distance between peaks indicates a more stable CP. Baratto et al. [17] showed that the discriminative power of these two sway density parameters, together with the mean speed, is greater than that of other global parameters (for instance, range of the CP and RMS values) to distinguish among sensory and pathological conditions in the general framework of balance control. All computations were performed using Matlab 7.0 (The MathWorks, Natick, MA).

Table 2 Postural parameters of the subjects with vision and without vision Postural parameters

Eyes open

Eyes closed

Range ML (cm) Range AP (cm) RMS ML (cm) RMS AP (cm) RMS ML velocity (cm/s) RMS AP velocity (cm/s) CP speed (cm/s) Sway area (cm2) Mean distance (mm) Mean peak (s)

0.95  0.4 [0.26–2.48] 1.75  0.7 [0.57–5.42] 0.21  0.09 [0.06–0.52] 0.39  0.14 [0.13–1.07] 0.42  0.22 [0.13–1.62]

1.07  0.6 [0.31–3.32] 2.19  0.8 [0.69–4.81] 0.23  0.10 [0.07–0.55] 0.46  0.16 [0.16–0.93] 0.51  0.25 [0.16–1.42]

0.98  0.37 [0.36–2.30]

1.48  0.71 [0.44–4.39]

0.91  0.3 [0.37–1.94] 1.03  0.9 [0.09–6.35] 2.23  0.7 [0.95–4.88]

1.27  0.5 [0.40–2.51] 1.43  1.4 [0.14–7.78] 3.39  1.5 [1.40–7.78]

3.88  1.9 [1.11–8.57]

2.27  1.3 [0.53–6.11]

Values are means  S.D. and [range] of 59 subjects.

2.3. Statistical analyses A forward stepwise multiple regression analysis was used to determine if body weight could predict balance stability. In previous studies, age, body height and foot length also have been considered as predictors of postural stability [12,16,21,22]. For this reason, these variables were entered into the regression model. The Kolmogorov–Smirnov test was used to verify that all variables were normally distributed. Correlation coefficients were computed between all independent variables (predictors) to determine the level of colinearity. Body weight did not correlate with age and height ( p > 0.05). A small but significant correlation was observed between body weight and foot length (r2 = 0.13, p < 0.01). Other anthropometric parameters were also examined (body mass index, hip and waist circumference) but not considered for the model because they highly correlated with body weight (r2 ranging from 0.90 to 0.95 for the variables mentioned above). Variables were entered in the model when the F value was greater than 1.0. Mathematical weightenings of the explanatory variables in the equation (b) and estimated precision of the coefficient (S.E. b) were examined and presented. They give an indication of the relative importance of the variables entered into the model in explaining variance observed in balance control. All results were considered to be significant at the 5% critical level ( p < 0.05). Backward stepwise analyses also were conducted. The significant predictors were the same than those obtained with the forward approach and the percentage of explained variance varied by less than 1% for all analyses. For this reason, only results for the forward approach are presented. Statistica Software 7.0 (Statsoft, Inc., Tulsa, OK) was used for all analyses.

3. Results Mean values for all balance control parameters (with and without vision) analyzed are shown in Table 2. Forward stepwise multiple regression analyses were conducted for all

CP variables. Table 3 presents results of the stepwise multiple regression analyses for the CP speed with and without vision. With vision, the stepwise multiple regression indicated that, among all parameters, body weight was the only significant predictive factor of CP speed ( p < 0.001). It accounted for 52% of the variance of CP speed. As indicated in Table 3, adding age contributed to explain an additional fraction of 3% of the variance of CP speed. In this case, the statistical contribution of body height and foot length did not reach the statistical inclusion criteria. The final model explained 55% of the variance of CP speed (r = 0.74). Without vision, body weight and age were identified as significant predictive factors of CP speed ( p < 0.001 and p < 0.01, respectively). Moreover, the stepwise multiple regression revealed that when body weight was entered into the model at the first step, it accounted for 54% of the variance of CP speed without vision ( p < 0.001). As indicated in Table 3, adding age at the second step and body height at the third step added a further 8% and 1% to explain the variance of CP speed. Foot length did not satisfy the statistical inclusion criteria. The final model explained 63% of the variance of CP speed (r = 0.79). Fig. 1 presents CP speed as a function of body weight with vision (1A) and without vision (1B). The striking observation is that, both with and without vision, CP speed increases linearly when body weight increases, indicating that larger body weight reduces postural balance stability. Stepwise multiple regression analyses were performed for each posturographic parameter separately (Table 4). The same general picture emerged with body weight being the factor always explaining the largest percentage of the observed variance (adjusted r2 values ranging from 0.08 to 0.47 with vision and 0.16 to 0.54 without vision). For most regressions, age contributed to a small portion of the variance. Body height and foot length had limited explanatory power. When weight and height were replaced by the interactive factor BMI, the percentage of explained variance decreased (because of colinearity, weight and BMI

O. Hue et al. / Gait & Posture 26 (2007) 32–38

35

Table 3 Hierarchical stepwise multiple regression examining the effect of age, body height, foot length and body weight on the postural parameter CP speed with vision and without vision (n = 59) b CP speed (with vision) Body weight Age Final solution of the model (body weight + age) CP speed (without vision) Body weight Age Body height Final solution of the model (body weight + age + body height)

S.E. b t

95% to 95% CI b p-Level r

r2

Adjusted r2 F

p-Level

0.69 0.09 0.16 0.09

t(56) = 7.40 t(56) = 1.68

0.49 to 0.89 0.06 to 0.34

<0.001 >0.09

0.72 0.52 0.52 0.74 0.55 0.53 0.74 0.55 0.53

F(1, 57) = 62.84 <0.001 F(2, 56) = 33.84 <0.001 F(2, 56) = 33.84 <0.001

0.68 0.09 0.27 0.09 0.09 0.08

t(55) = 7.94 t(55) = 3.12 t(55) = 1.04

0.48 to 0.84 0.06 to 0.43 0.33 to 0.08

<0.001 <0.01 >0.30

0.74 0.79 0.79 0.79

F(1, F(2, F(3, F(3,

0.54 0.62 0.63 0.63

0.54 0.60 0.61 0.61

57) = 68.17 56) = 45.69 55) = 30.85 55) = 30.85

<0.001 <0.001 <0.001 <0.001

b: mathematical weightenings of the explanatory variables in the equation. S.E. b: estimated precision of the coefficient. t: value of the test statistic from the t distribution. 95% to 95% CI: 95% confidence intervals for the coefficient. The correlation coefficient (r), the multiple correlation coefficient (r2), the adjusted multiple correlation coefficient (adjusted r2) and the p-level are reported after each step.

cannot be entered simultaneously in a model). It is noteworthy that the percentage of explained variance was always more important for the no-vision than the vision condition. Also, for both visual conditions, the percentage of

variance explained by the model was always greater when the dependent variable was CP speed.

4. Discussion

Fig. 1. Relationship between body weight and CP speed with vision (A) and without vision (B).

The main objective of the study was to determine whether body weight could predict balance stability. With and without vision, body weight alone contributed to more than 50% of the variance observed for CP speed when controlling for age, body height and foot length. Without vision, the final statistical model predicted 63% of the variance observed for CP speed (55% with vision). Table 4 points to an interesting observation. With and without vision, body weight explained considerable variance for mean speed, mean peak and mean distance. It is noteworthy that Baratto et al. [17] also pointed out the sensitivity of these three parameters (CP speed, mean distance and mean peak) for distinguishing various sensory and pathological conditions. Mean speed is often considered to represent an overall amount of activity necessary to maintain stability. Mean peak corresponds to time instants in which the ankle torque and the associated motor commands are relatively stable and mean distance represents the distance between one relative stable region to another one. With the increase of body weight, the peaks decreased and the distance between stable regions increased significantly. Altogether, these observations suggest that with a heavier weight the balance control system was less sensitive to regulate body sway oscillations. Overall, the data points out to an important limitation factor for obese persons as a faster CP speed (and decreased stability) has been repeatedly associated with an increased risk of falling [16,23]. Balance control depends on the integration of information from all sensorimotor systems (visual, vestibular, proprioceptive). A small sway deviation from a perfect vertical position leads to a torque due to gravity that moves and accelerates the body further away from the upright neutral

36

O. Hue et al. / Gait & Posture 26 (2007) 32–38

Table 4 Summary of the stepwise multiple regression analysis examining the effect of body weight, age, body height and foot length on postural parameters with and without vision (n = 59) Postural parameters

With vision

Postural parameters

Final solution of the model

r2

Adjusted r2

p-Level

CP speed

(1) Body weight***, (2) age

0.55

0.53

<0.001

CP speed

Mean distance

(1) Body weight***

0.48

0.47

<0.001

RMS VCOP AP

RMS VCOP AP

(1) Body weight***, (2) age***, (3) body height

0.47

0.44

<0.001

Mean distance

Mean peak

(1) Body weight***, (2) age

0.42

0.40

<0.001

Mean peak

RMS ML

(1) Body weight*, (2) foot length

0.19

0.16

<0.01

Range AP

Range ML

(1) Body weight, (2) foot length

0.15

0.12

<0.05

RMS AP

RMS VCOP ML

(1) Age, (2) body weight

0.13

0.10

<0.05

RMS ML

Sway area

(1) (2) (3) (1) (2) (1)

0.13

0.09

0.05

Sway area

0.12

0.09

<0.05

RMS VCOP ML

0.09

0.08

<0.05

Range ML

Range AP RMS AP

Body weight, foot length, body height Body weight, age Body weight*

Without vision Final solution of the model (1) (2) (3) (1) (2) (3) (1) (2) (3) (4) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2)

Body weight***, age**, body height Body weight***, age***, body height Body weight***, age, body height, foot length Body weight***, age*, body height Body weight***, age**, body height Body weight***, age*, body height Body weight**, age, foot length Body weight**, age*

(1) (2) (1) (2)

Body weight**, age Body weight*, age

r2

Adjusted r2

p-Level

0.63

0.61

<0.001

0.56

0.54

<0.001

0.54

0.51

<0.01

0.49

0.46

<0.001

0.43

0.40

<0.05

0.40

0.37

<0.001

0.26

0.22

<0.001

0.25

0.22

<0.001

0.19

0.16

<0.01

0.19

0.16

<0.01

The parameters are ordered by decreasing r2. The correlation coefficient (r), the multiple correlation coefficient (r2), the adjusted multiple correlation coefficient (adjusted r2) and the p-level are reported. Significant explanatory factor at the level: *p < 0.05, **p < 0.01 and ***p < 0.001.

position. A corrective torque exerted by the feet counteracts this destabilizing torque. A widely held view is that the corrective torque is generated through the action of a feedback control system. There are at least two reasons for explaining why balance stability is strongly correlated and predicted by body weight. A first explanation relates to the contribution of foot mechanoreceptors and cutaneous sensation for balance control [24–27]. Bensmaı¨a et al. [28] recently showed that prolonged suprathresold vibratory stimulation can produce a desensitization of mechanoreceptor afferents. Their result is important because it suggests the discriminatory power of the mechanoreceptors can be affected by a constant stimulus. Overweight is a likely candidate to reduce the sensitivity of mechanoreceptors. Several recent experiments for which plantar contact areas and pressure were measured for obese persons point to this suggestion. When compared to nonobese persons, obese persons generally show larger plantar contact areas and greater mean pressure values for most anatomical landmarks tested [29] and significant increases in pressure under the heel, mid-foot and metatarsal heads [30]. Hence, the greater pressure values and larger contact areas for the obese persons may reduce the quality and/or quantity of the sensory information arising from the plantar mechan-

oreceptors. It is likely that plantar mechanoreceptors participate to the feedback control system regulating body sway oscillations as they are related to different parameters of ground reaction force which are indirectly related to CP displacements [31]. This suggestion is reinforced by recent results from our group [14] where we showed that a weight loss intervention in obese and morbid obese individuals yielded an increased postural instability (decreasing range and speed of CP with weight loss). There are other possibilities for explaining the strong relationship between balance stability and body weight. When standing upright, the human body is often compared to an inverted pendulum system rotating around the ankle joint. The center of mass located closer to the anterior edge of the base of support, due to extra abdominal mass, presumably leads to an increased ankle torque necessary to maintain balance [8]. Greater ankle torque could add more noise in the feedback control system as greater muscle force is related to greater motor variability [32,33]. Therefore, it is likely that the central command, allowing body sway regulation, is not adapted due to reduced capability of the mechanoreceptors to accurately signal the position of the CP and to greater motor variability.

O. Hue et al. / Gait & Posture 26 (2007) 32–38

The present study shows that an increase in body weight correlates with a greater balance instability. The prevalence of obesity is increasing in all age groups including older persons [34]. This may put at higher risk the older obese person since balance instability is one of the most important risk factor leading to fall in this population [35–37]. More than one third of persons 65 years of age or older fall each year [36] and about 10% of these falls lead to serious injury [35,38]. This suggestion may appear conflicting with several observations arising from epidemiological studies suggesting that a low BMI is a risk factor of fall [39–42]. In several studies, it was proposed that BMI could provide: (i) skeletal loading, which may cause a compensatory increase in bone mass [43–45], and (ii) padding that protects against fractures during falls [46]. We do not want to argue against these potential benefits from a high BMI. These studies, however, do not provide an indication of the risk of falling related to obesity. The benefit from a high BMI do not prevent people from falling but it may attenuate the health related consequences of a fall for frail persons. In support of this suggestion, there are some studies that have reported a significant relationship between obesity and risk of falling [47,48]. Also, a recent epidemiological study by Nguyen et al. [49] shows that, after adjusting for bone mineral density or body weight, the abdominal fat-fracture association previously reported is no longer statistically significant. Clearly, continuing studies should assess the specific risk of falling in obese subjects rather than extrapolating risk for falling from reported injuries or fractures. In conclusion, body weight predicts variation in balance stability which is an essential prerequisite in daily life activities. As recently argued by Wearing et al. [6], there is an urgent need to examine objectively the actual impact of obesity on balance control and on the diverse range of activities of daily living.

Acknowledgments Supported by a NSERC collaborative health grant and NSERC discovery grants. Thanks are expressed to Franc¸ois Be´gin for his help in data collection.

References [1] Kopelman PG. Obesity as a medical problem. Nature 2000;404:635– 43. [2] Janssen I, Katzmarzyk PT, Ross R. Body mass index, waist circumference, and health risk: evidence in support of current National Institutes of Health guidelines. Arch Intern Med 2002;162:2074–9. [3] Bray GA. Medical consequences of obesity. J Clin Endocrinol Metab 2004;89:2583–9. [4] Rodacki AL, Fowler NE, Provensi CL, Rodacki Cde L, Dezan VH. Body mass as a factor in stature change. Clin Biomech 2005;20:799– 805. [5] Fabris de Souza SA, Faintuch J, Valezi AC, et al. Postural changes in morbidly obese patients. Obest Surg 2005;15:1013–6.

37

[6] Wearing SC, Hennig EM, Byrne NM, Steele JR, Hills AP. The biomechanics of restricted movement in adult obesity. Obest Rev 2006;7:13–24. [7] Lord SR, Sturnieks DL. The physiology of falling: assessment and prevention strategies for older people. J Sci Med Sport 2005;8:35–42. [8] Corbeil P, Simoneau M, Rancourt D, Tremblay A, Teasdale N. Increased risk for falling associated with obesity: mathematical modeling of postural control. IEEE Trans Neural Syst Rehabil Eng 2001;9:126–36. [9] Goulding A, Jones IE, Taylor RW, Piggot JM, Taylor D. Dynamic and static tests of balance and postural sway in boys: effects of previous wrist bone fractures and high adiposity. Gait Posture 2003;17: 136–41. [10] McGraw B, McClenaghan BA, Williams HG, Dickerson J, Ward DS. Gait and postural stability in obese and nonobese prepubertal boys. Arch Phys Med Rehabil 2000;81:484–9. [11] Petti S, Cairella G, Tarsitani G. Childhood obesity: a risk factor for traumatic injuries to anterior teeth. Endod Dent Traumatol 1997; 13:285–8. [12] Chiari L, Rocchi L, Cappello A. Stabilometric parameters are affected by anthropometry and foot placement. Clin Biomech 2002;17: 666–77. [13] McIlroy WE, Maki BE. Preferred placement of the feet during quiet stance: development of a standardized foot placement for balance testing. Clin Biomech 1997;12:66–70. [14] Teasdale N, Hue O, Marcotte J, Berrigan F, Simoneau M, Dore´ J, et al. Reducing weight increases postural stability in obese and morbid obese men. Int J Obest 2007;31:153–60. [15] Geurts ACH, Nienhuis B, Mulder TW. Intrasubject variability of selected force-platform parameters in the quantification of postural control. Arch Phys Med Rehabil 1993;74:1144–50. [16] Maki BE, Holliday PJ, Topper AK. A prospective study of postural balance and risk of falling in an ambulatory and independent elderly population. J Gerontol 1994;49:72–84. [17] Baratto L, Morasso PG, Re C, Spada G. A new look at posturographic analysis in the clinical context: sway-density versus other parameterization techniques. Motor Control 2002;6:246–70. [18] Raymakers JA, Samson MM, Verhaar HJ. The assessment of body sway and the choice of the stability parameter(s). Gait Posture 2005; 21:48–58. [19] Riley MA, Baker AA, Schmit JM, Weaver E. Effects of visual and auditory short-term memory tasks on the spatiotemporal dynamics and variability of postural sway. J Motor Behav 2005;37:311–24. [20] Jacono M, Casadio M, Morasso PG, Sanguineti V. The sway-density curve and the underlying postural stabilization process. Motor Control 2004;8:292–311. [21] Balogun JA, Akindele KA, Nihinlola JO, Marzouk DK. Age-related changes in balance performance. Disabil Rehabil 1994;16:58–62. [22] Kejonen P, Kauranen K, Vanharanta H. The relationship between anthropometric factors and body-balancing movements in postural balance. Arch Phys Med Rehabil 2003;84:17–22. [23] Fernie GR, Gryfe CI, Holliday PJ, Llewellyn A. The relationship of postural sway in standing to the incidence of falls in geriatric subjects. Age Ageing 1982;11:11–6. [24] Maki BE, Perry SD, Norrie RG, McIlroy WE. Effect of facilitation of sensation from plantar foot-surface boundaries on postural stabilization in young and older adults. J Gerontol A Biol Sci Med Sci 1999; 54:281–7. [25] Perry SD, McIlroy WE, Maki BE. The role of plantar cutaneous mechanoreceptors in the control of compensatory stepping reactions evoked by unpredictable, multi-directional perturbation. Brain Res 2000;877:401–6. [26] Kavounoudias A, Roll R, Roll JP. Foot sole and ankle muscle inputs contribute jointly to human erect posture regulation. J Physiol 2001; 532:869–78. [27] Meyer PF, Oddsson LI, De Luca CJ. The role of plantar cutaneous sensation in unperturbed stance. Exp Brain Res 2004;156:505–12.

38

O. Hue et al. / Gait & Posture 26 (2007) 32–38

[28] Bensmaı¨a SJ, Leung YY, Hsiao SS, Johnson KO. Vibratory adaptation of cutaneous mechanoreceptive afferents. J Neurophysiol 2005;94: 3023–36. [29] Birtane M, Tuna H. The evaluation of plantar pressure distribution in obese and non-obese adults. Clin Biomech 2004;19:1055–9. [30] Hills AP, Hennig EM, McDonald M, Bar-Or O. Plantar pressure differences between obese and non-obese adults: a biomechanical analysis. Int J Obest Relat Metab Disord 2001;25:1674–9. [31] Morasso PG, Schieppati M. Can muscle stiffness alone stabilize upright standing? J Neurophysiol 1999;82:1622–6. [32] Jones KE, Hamilton AF, Wolpert DM. Sources of signal-dependent noise during isometric force production. J Neurophysiol 2002;88: 1533–44. [33] Sherwood DE, Schmidt RA. The relationship between force and force variability in minimal and near-maximal static and dynamic contractions. J Motor Behav 1980;12:75–89. [34] Mokdad AH, Bowman BA, Ford ES, et al. The continuing epidemics of obesity and diabetes in the United States. JAMA 2001;286:1195– 200. [35] Tinetti ME. Clinical practice preventing falls in elderly persons. N Engl J Med 2003;348:42–9. [36] Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319: 1701–7. [37] Kannus P, Sievanen H, Palvanen M, Jarvinen T, Parkkari J. Prevention of falls and consequent injuries in elderly people. Lancet 2005; 366:1885–93. [38] Nevitt MC, Cummings SR. Type of fall and risk of hip and wrist fractures: the study of osteoporotic fractures. J Am Geriatr Soc 1993; 41:1226–34.

[39] Tinetti ME, Doucette JT, Claus EB. The contribution of predisposing and situational risk factors to serious fall injuries. J Am Geriatr Soc 1995;43:1207–13. [40] Koski K, Luukinen H, Laippala P, Kivela SL. Risk factors for major injurious falls among the home-dwelling elderly by functional abilities. A prospective population-based study. Gerontology 1998;44:232–8. [41] Young Y, Myers AH, Provenzano G. Factors associated with time to first hip fracture. J Aging Health 2001;13:511–26. [42] Willig R, Luukinen H, Jalovaara P. Factors related to occurrence of hip fracture during a fall on the hip. Public Health 2003;117:25–30. [43] Nevitt MC, Cummings SR. Type of fall and risk of hip and wrist fractures: the study of osteoporotic fractures. J Am Geriatr Soc 1993;41:1226–34. [44] Langlois JA, Visser M, Davidovic LS, et al. Hip fracture risk in older white men is associated with change in body weight from age 50 years to old age. Arch Intern Med 1998;158:990–6. [45] Grisso JA, Kelsey JL, Strom BL, et al. Risk factors for falls as a cause of hip fracture in women. The Northeast Hip Fracture Study Group. N Engl J Med 1991;324:1326–31. [46] Owusu W, Willett W, Ascherio A, et al. Body anthropometry and the risk of hip and wrist fractures in men: results from a prospective study. Obest Res 1998;6:12–9. [47] Bergland A, Pettersen AM, Laake K. Functional status among elderly Norwegian fallers living at home. Physiother Res Int 2000;5:33–45. [48] Wallace C, Reiber GE, LeMaster J, et al. Incidence of falls, risk factors for falls, and fall-related fractures in individuals with diabetes and a prior foot ulcer. Diabetes Care 2002;25:1983–6. [49] Nguyen ND, Pongchaiyakul C, Center JR, Eisman JA, Nguyen TV. Identification of high-risk individuals for hip fracture: a 14-year prospective study. J Bone Miner Res 2005;20:1921–8.