Journal of Bodywork & Movement Therapies (2015) 19, 380e388
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/jbmt
PREVENTION & REHABILITATION: SINGLE CASE STUDY
Can osteopathic manipulative treatment modify the posture in elderly people? e A single-case study PREVENTION & REHABILITATION: SINGLE CASE STUDY
´neuc, G. Guihard* F. Pellerin, P. Guihe Laboratoire de Neurophysiologie Expe´rimentale, Faculte´ de Me´decine, Universite´ de Nantes, Place Ricordeau, 44035 Nantes Cedex, France Received 3 April 2014; received in revised form 22 June 2014; accepted 26 June 2014
KEYWORDS Postural control; Aging; Fast Fourier transform; Centre of foot pressure; Falling; Sway; Manual therapy
Summary In this research, we have studied the consequences of three consecutive osteopathic manipulative sessions (OMS) on postural control by using a single-case research (SCR) design. The patient was a 77 years old woman complaining of altered balance and low-back pain. OMS were delivered by a single practitioner. The pain level was self-rated by using a visual Borg scale. The posture was monitored on a force platform. Postural parameters were deduced from the analysis of the centre of foot pressure (CoP) displacement. The statistical significance of the observed differences was established by using an SCR-related effect size indicator (i.e. Taunovlap). Our results indicate that OMS decrease the patient’s pain, modify CoP mean position and decreased the length and velocity of the CoP displacement. Furthermore, modifications of the body oscillations were observed after OMS. This work indicates that OMS can improve body balance and that SCR allows the objective evaluation of the consequences of OMS. ª 2014 Elsevier Ltd. All rights reserved.
Introduction Randomized controlled trials (RCT) have been successfully used for the evaluation of osteopathic and chiropractic manipulations, leading to recognized results and conclusions (Hodge and Downey, 2011; Posadzki and Edzard, 2011; Chaibi and Russell, 2012; Cerritelli et al., 2013). However, * Corresponding author. Tel.: þ33 (0)6 69 50 99 74. E-mail address: [email protected] (G. Guihard). http://dx.doi.org/10.1016/j.jbmt.2014.06.002 1360-8592/ª 2014 Elsevier Ltd. All rights reserved.
several practitioners have pointed to the inability of manual therapies to fit with RCT methodology, claiming that each patient is unique and that manipulative treatment has to be patient-specific (Littlewood, 2011; Milanese, 2011). Furthermore, the integration of a placebo manipulation in the evaluation procedure is a controversial (Mein et al., 2001; Noll et al., 2004). Thus, the
evaluation of osteopathic manipulations is still insufficiently documented and presents insufficient objective arguments (Brantingham et al., 2011; Ja ¨kel and von Hauenschild, 2011; Pizzolorusso et al., 2011). The single-case research (SCR) is a well-accepted method for physiological, psychological or behavioural characterization in single healthy subject. SCR principles can be obtained from various review articles (Nock et al., 2007; Rizvi and Nock, 2008; Romeiser Logan et al., 2008). The use of SCR is recommended when the constitution of research groups is unlikely, due to the very low occurrence of pathology, as in autoimmune manifestations (Takagi et al., 2011). It is also recommended when the recruitment of patients presenting a similar pathological development is difficult, as in autism (Lin et al., 2012) or as in traumatic brain injury (Hsieh, 2008). SCR can also be employed when the therapeutical protocol is customized to the status of the patient, as in physical rehabilitation (Hsieh, 2008; Klein et al., 2011; Lima-Gregio et al., 2011; Ross and Elliott, 2011; Arazpour et al 2013, Peter et al., 2012; Schubert et al., 2012; Smania et al., 2012). SCR has also been considered as a possible method for the scientific evaluation of manual therapies (Wolery and Harris, 1982; Sanders, 2003) and several reports argue for this (Maricar et al., 2009; Puentedura et al., 2010; Treleaven, 2010). Recently, SCR has been used to characterize the benefits of chiropractic manipulations on neck pain and headache (Montan ˜ez-Aguilera et al., 2010; Chaibi and Tuchin, 2011). Conventional effect size cannot be calculated to estimate differences generated by a treatment vs. baseline in a single subject. Indeed, SCR avoids the use of inferential statistical tests. However, several works have proposed alternative methods for differences evaluation by using non-parametric distribution-free models (Parker and Vannest, 2009; Wolery et al., 2010; Parker et al., 2011a). Several indices can be determined from data pairwise comparison with statistical arguments, making these more convenient for estimating treatment-related changes (Parker and Vannest, 2009; Parker et al., 2011a and b). In the present work, we hypothesized that the consequences of osteopathic manipulative treatment can be evaluated through a SCR analysis. We recruited an elderly woman who complained for body instability and low back pain. The patient was treated by three consecutive osteopathic manipulative sessions. Primary and secondary outcomes consisted in monitoring the displacement of the patient’s centre of foot pressure (CoP) onto a force platform and the pain level before and after OMS.
Material and methods Patient characteristics The patient (recruited in December 2011 at IdHEO Osteopathic School, Nantes, France) was a 77 years old Caucasian woman (height Z 1.60 m, weight Z 70 kg, BMI Z 27.3). Before the recruitment, the patient had never consulted an osteopathic practitioner. The patient declared to be a housewife and reported four pregnancies in 1964, 1966, 1967 and 1970. The first one was complicated by coccyx
381
dislocation at the delivery; the last three proceeded without any complication. No perineum rehabilitation was required between pregnancy episodes. The patient did not exhibit any acute pathologies but was being pharmacologically treated for moderate arterial hypertension, vascular thrombosis, hyperlipidaemia, vitamin D deficiency and diffuse back pain episodes (over the previous six years). Treadmill walking did not reveal any cardiac dysfunctions. The patient’s husband has become hemiplegic after a cerebro-vascular thrombosis in 2005. As a consequence of this, the autonomy of the patient’s husband for feeding and personal hygiene was strongly diminished. The patient declared a strong physical involvement in the daily care of her husband. Furthermore, the patient reported two falling episodes during summer 2011. A NMR imagery of the head-neck region did not reveal any cerebral or spinal alteration. The absence of neuro-anatomical symptoms was noted. At the time of the recruitment, the patient reported low back pain in the area of both sacroiliac joints that may have been related to physical efforts necessary for the daily care of her husband. Moreover, the patient complained of a general feeling of postural imbalance, not associated with vertigo. The patient had not undergone previous postural investigations before this study. The patient was informed about the aim and the conditions of this study. She gave her informed consent according to the principles of the Helsinki declaration. The general conditions of the study were approved by IdHEO Nantes ethical committee composed by three senior osteopathic manipulators, two scientists (PhD) and one medical doctor (MD, PhD).
Monitoring of the postural control The position of the centre of foot pressure (CoP) was determined by using the WinPosturo AFP 40/16 force platform and the WinPosture NV software (Medicapteurs France SAS, Balma, France). Three strain gauges form the three tops of a virtual equilateral triangle on the platform. The patient’s feet were adjusted onto the platform by a removable plexiglass device so that the measures were realized reproducibly. In these recording conditions, the CoP position ideally corresponds to the triangle barycentre, the coordinates of which are 0 mm in both antero-posterior and medio-lateral directions. The antero-posterior and medio-lateral values of the CoP position of our patient were given as a function of this 0 mm reference. Left or right lateral shifts of the CoP position correspond to negative or positive medio-lateral values, while a front or rear shift corresponds to positive or negative values in the antero posterior direction. The patient was asked to focus on a target located on a vertical wall, at 90 cm from her eyes. Furthermore, the patient was asked not to clench her jaw and to stand as still as possible during the time of recording (Zok et al., 2008). The measures were repeated 5 times before and after each OMT session. Two consecutive measures were separated by a rest episode (duration: 30 s) during which the patient remained seated. For one measure, the CoP displacement was acquired for 51.2 s (acquisition
PREVENTION & REHABILITATION: SINGLE CASE STUDY
Can osteopathic manipulative treatment modify elderly posture
382 frequency Z 40 Hz) (Sasaki et al., 2002). This acquisition time is imposed by the manufacturer. The signal was amplified, low-pass filtered (cut off frequency Z 10 Hz), digitized (sampling frequency Z 20 Hz) through a 16 bits analogic/digital converter, and finally stored onto a personal computer for further analysis.
PREVENTION & REHABILITATION: SINGLE CASE STUDY
Outcome measurements Postural parameters From each episode of CoP monitoring, seven postural parameters were automatically calculated from the displacement of the CoP by WinPosture NV software. These corresponded to the mean medio-lateral CoP position (MLP in mm) and the mean antero-posterior CoP position (APP in mm), the surface area (S in mm2) of the CoP displacement, the total length (L in mm), the medio-lateral length (MLL in mm) and the antero-posterior length (APL in mm) of the CoP path, and the mean CoP velocity (V in mm$s1) during the CoP displacement. The displacement of CoP corresponds to body oscillations that are related to particular frequencies. The Fast Fourier transform (FFT) is a conventional mathematical procedure for the decomposition of an oscillating signal into a frequency spectrum. It allows the frequency analysis of a time-related physiological function including the displacement of CoP (Pichon et al., 2006; Cabeza-Ruiz et al., 2011). FFT calculates the energy (in V2 Hz1) associated with each frequency of the spectrum. The summation of all frequency-related energies corresponds to the power spectrum total energy (PSTE). A treatment can induce some modifications of body oscillations through modifications of the PSTE. One can analyse these modifications through the total energy (rough information) or through each frequency-related peak of energy (extremely precise information). According to previous works devoted to frequency analysis of postural control, the total power spectrum energy is distributed in three frequency bands depending on the type of somatic regulation: very low frequencies (0e0.5 Hz) correspond to the action of the sensory systems (visual and vestibular), low frequencies (0.5e2 Hz) are related to the regulation of the cerebellum, and medium frequencies (>2 Hz) reflect proprioceptive regulation (Dichgans et al., 1976; Njiokiktjien et al., 1978; Nagy et al., 2004; Bizid et al., 2009; Cabeza-Ruiz et al., 2011). In our conditions, the oscillations of the patient were automatically analysed for the three frequency bands: very low frequencies (VLF: from 0 to 0.5 Hz), low frequencies (LF: from 0.5 to 2 Hz), and medium frequencies (MF: from 2 to 5 Hz). The contribution of each frequency domain to the overall body oscillation was expressed as a percentage of PSTE. This was done for oscillations in both ML and AP directions.
Osteopathic protocol A strategy including three osteopathic manipulative sessions (OMS) was defined. The overall procedure consisted in an A-B-A-C-A-D protocol, with A corresponding to baseline measurements before OMS and B, C, and D correspond to
F. Pellerin et al. the outcome measures after first, second and third OMS, respectively. Each OMS lasted 30 min. All OMS were performed by the same French trained osteopathic practitioner (two years’ experience after an initial 5 years osteopathic training). One week separated two consecutive OMS, during which the patient was asked not to modify her way of life. First OMS (OMS01) From osteopathic tests and examination, the following points were observed : a limited mobility of the right coxofemoral joint during external and internal rotation; a restriction of the iliac bone’s posterior mobility; the coccyx tilted in the lower left direction; the right psoas muscle remained contracted and the right diaphragm dome stayed in a low position during expiration. The OMS consisted in: i) the coxofemoral and sacroiliac joints mobilization by a general osteopathic treatment of the lower limbs (the patient remained supine) (Nicholas and Nicholas, 2008); ii) the diaphragm was released by helping the diaphragm upward movements during respiration, by inserting the ulnar side of the practitioner hand beneath the patient chest, and by pushing the right diaphragm dome in a superior direction (the patient remained supine) (Nicholas and Nicholas, 2008); iii) the restoration of coccygeal mobility by applying appropriate mobilising pressure during expiration (the patient was prone) (Nicholas and Nicholas, 2008). Second OMS (OMS02) From osteopathic assessment, it was deduced there existed limited mobility of the right talus, the right external malleolus and an externally translated right patella. The right knee appeared compressed. The sacrum tilted inferiorly and the right piriformis muscle was hypertonic. The osteopathic treatment consisted in: i) a general osteopathic treatment of the right foot allowing a muscular release in the right ankle region (the patient remained supine) (Nicholas and Nicholas, 2008); ii) structural and high velocity low amplitude manipulations were performed, involving the right talus (the patient remaining in decubitus) and the right fibula (the patient lay on the left side) (Nicholas and Nicholas, 2008); iii) inhibition of the piriformis muscle was carried out (the patient remained prone) (Nicholas and Nicholas, 2008); iv) the increase of the sacrum mobility (the patient remaining in procubitus) (Nicholas and Nicholas, 2008); v) decompression of the right knee (the patient sitting on the manipulation table with the table edge inserted in his popliteal fossa) ((Nicholas and Nicholas, 2008). Third OMS (OMS03) From osteopathic tests, a lumbar restriction at the Th12-L1 and L5-S1 levels were noted, in association with a hypomobility of the sternum and the right anterior region of the chest. The right diaphragm dome stayed in a low position during exhalation, as already observed. The osteopathic treatment consisted in: i) a general osteopathic treatment of the chest (Nicholas and Nicholas, 2008); ii) mobilization of the sternum (Nicholas and Nicholas, 2008); iii) relaxation of the right diaphragm dome (Nicholas and Nicholas, 2008); iv) a bilateral general
Can osteopathic manipulative treatment modify elderly posture
Data analysis We calculated means, standard deviations (sd) and 95% confident intervals (CI95) from our data by using Sigma Plot 12.0 (Jandel Scientific). In SCR, the comparison of data sets by inferential statistics cannot be done since the independence of the values used for mean calculation is not established. Here, a data pairwise comparison was used to analyse any overlap between data sets obtained before and after OMS. We proceeded according to the method presented by Parker et al. (2011a). Actually, the results of the 5 measures obtained before and after each OMS were represented as a function of the number of the measure in a time-forward direction. Each result of the previous session was compared to each result of the subsequent session. Thus, 25 data pairs were generated for each OMS. For each data pair, the result of the subsequent session could be higher, lower, or equivalent to the result of the earlier session. Thus, the numbers of improving, declining and identical data pairs were counted for each OMS. From these, Taunovlap could be calculated as :
negative value corresponded to a decrease of the parameter value. Taunovlap value is not an estimation of the statistical significance of a difference between before and after data sets. As pointed by Parker et al. (2011a,b) Taunovlap can also be obtained from a ManneWhitney U test, from which a p value can be deduced. The different values for Taunovlap and p were calculated by using an automatic web-based calculator located at http://www.singlecaseresearch.org. We considered that p < 0.05 indicated that the difference between before and after data sets was significant.
Results Postural parameters We calculated the values of the seven postural parameters (MLP, APP, S, L, MLL, APL and V) for five CoP displacements recorded before and after each OMS. The values for means, standard deviations and CI95 are presented in Table 1. Medio-lateral and antero-posterior positions of CoP At the onset of the study, the mean values for MLP and APP were 14.0 mm and 42.6 mm, respectively (Table 1). Following OMS01, MLP and APP mean values were
ðnumber of improving pairsÞ ðnumber of declining pairsÞ Taunovlap Z100$ total number of pairs
A positive value for Taunovlap was expected when the number of improving pairs exceeded the number of declining pairs. If Taunovlap was 100%, this corresponded to no overlapping between the two data sets. Thus, a positive value indicated that the treatment has promoted a positive evolution of the measured parameter. Alternatively a
12.5 mm and 42.4 mm, respectively (Table 2). From data pairwise comparison after vs. before conditions, we calculated Taunovlap Z 44% for MLP and Taunovlap Z 28% (Fig. 1) for first OMS. Considering p values (Fig. 1), we found that these differences were not significant. Similar calculations and comparisons were made for MLP and APP values
Table 1 The values of the different postural parameters (MLP, APP, S, V, L, MLL and APL) were calculated from the CoP displacement recorded before and after each OMT session. Mean, sd, and 95% confident interval (CI95) were determined from 5 recordings obtained before and after each OMT session. OMS01
MLP (mm) APP (mm) S (mm2) V (mm$s1) L (mm) MLL (mm) APL (mm)
Mean CI95 Mean CI95 Mean CI95 Mean CI95 Mean CI95 Mean CI95 Mean CI95
(sd) (sd) (sd) (sd) (sd) (sd) (sd)
OMS02
OMS03
Before
After
Before
After
Before
After
13.4 (2.7) 16.8; 10.0 42.6 (1.2) 44.1; 41.1 127 (19) 104; 150 10.6 (1.1) 9.2; 12.0 497 (58) 425; 569 260 (36) 205; 305 383 (46) 326; 440
11.3 (2.0) 13.9; 8.7 42.4 (1.2) 43.9; 40.9 144 (14) 127; 162 9.2 (1.1) 7.8; 10.6 426 (56) 357; 495 237 (34) 195; 279 309 (48) 250; 368
10.4 (3.2) 14.3; 6.5 41.3 (4.1) 46.3; 36.3 113 (14) 95; 131 10.0 (1.0) 8.8; 11.2 470 (54) 403; 537 247 (36) 202; 292 356 (39) 307; 405
9.5 (3.6) 13.9; 5.1 38.6 (2.5) 41.6; 35.6 156 (26) 123; 189 9.3 (0.8) 8.3; 10.3 438 (45) 382; 494 237 (40) 188; 286 325 (29) 289; 361
9.5 (3.6) 13.8; 5.2 37.7 (1.5) 39.6; 35.8 120 (21) 93; 147 9.2 (0.6) 8.5; 9.9 432 (28) 397; 467 218 (24) 188; 248 330 (36) 285; 375
4.9 (1.5) 6.8; 3.0 37.5 (1.6) 39.5; 35.5 196 (41) 146; 246 8.8 (0.4) 8.3; 9.3 411 (23) 383; 439 232 (21) 206; 258 296 (13) 280; 312
PREVENTION & REHABILITATION: SINGLE CASE STUDY
osteopathic treatment in association with a stretching of the lumbo-sacral hinge (Nicholas and Nicholas, 2008).
383
384
F. Pellerin et al.
Table 2 The values of Taunovlap and p were calculated, as indicated in the methods section, by comparing data pairs obtained before and after each OMS and obtained before the first- and after the third OMS (01e03) for each postural parameter. Significant results are highlighted in bold. ns: not significant.
MLP APP S V L MLL
PREVENTION & REHABILITATION: SINGLE CASE STUDY
APL
Taunovlap p Taunovlap p Taunovlap p Taunovlap p Taunovlap p Taunovlap p Taunovlap p
OMS01
OMS02
OMS03
01 to 03
44% ns 28% ns 52% ns 60% ns 60% ns 36% ns 68% ns
12% ns 44% ns 100%
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/jbmt
PREVENTION & REHABILITATION: SINGLE CASE STUDY
Can osteopathic manipulative treatment modify the posture in elderly people? e A single-case study PREVENTION & REHABILITATION: SINGLE CASE STUDY
´neuc, G. Guihard* F. Pellerin, P. Guihe Laboratoire de Neurophysiologie Expe´rimentale, Faculte´ de Me´decine, Universite´ de Nantes, Place Ricordeau, 44035 Nantes Cedex, France Received 3 April 2014; received in revised form 22 June 2014; accepted 26 June 2014
KEYWORDS Postural control; Aging; Fast Fourier transform; Centre of foot pressure; Falling; Sway; Manual therapy
Summary In this research, we have studied the consequences of three consecutive osteopathic manipulative sessions (OMS) on postural control by using a single-case research (SCR) design. The patient was a 77 years old woman complaining of altered balance and low-back pain. OMS were delivered by a single practitioner. The pain level was self-rated by using a visual Borg scale. The posture was monitored on a force platform. Postural parameters were deduced from the analysis of the centre of foot pressure (CoP) displacement. The statistical significance of the observed differences was established by using an SCR-related effect size indicator (i.e. Taunovlap). Our results indicate that OMS decrease the patient’s pain, modify CoP mean position and decreased the length and velocity of the CoP displacement. Furthermore, modifications of the body oscillations were observed after OMS. This work indicates that OMS can improve body balance and that SCR allows the objective evaluation of the consequences of OMS. ª 2014 Elsevier Ltd. All rights reserved.
Introduction Randomized controlled trials (RCT) have been successfully used for the evaluation of osteopathic and chiropractic manipulations, leading to recognized results and conclusions (Hodge and Downey, 2011; Posadzki and Edzard, 2011; Chaibi and Russell, 2012; Cerritelli et al., 2013). However, * Corresponding author. Tel.: þ33 (0)6 69 50 99 74. E-mail address: [email protected] (G. Guihard). http://dx.doi.org/10.1016/j.jbmt.2014.06.002 1360-8592/ª 2014 Elsevier Ltd. All rights reserved.
several practitioners have pointed to the inability of manual therapies to fit with RCT methodology, claiming that each patient is unique and that manipulative treatment has to be patient-specific (Littlewood, 2011; Milanese, 2011). Furthermore, the integration of a placebo manipulation in the evaluation procedure is a controversial (Mein et al., 2001; Noll et al., 2004). Thus, the
evaluation of osteopathic manipulations is still insufficiently documented and presents insufficient objective arguments (Brantingham et al., 2011; Ja ¨kel and von Hauenschild, 2011; Pizzolorusso et al., 2011). The single-case research (SCR) is a well-accepted method for physiological, psychological or behavioural characterization in single healthy subject. SCR principles can be obtained from various review articles (Nock et al., 2007; Rizvi and Nock, 2008; Romeiser Logan et al., 2008). The use of SCR is recommended when the constitution of research groups is unlikely, due to the very low occurrence of pathology, as in autoimmune manifestations (Takagi et al., 2011). It is also recommended when the recruitment of patients presenting a similar pathological development is difficult, as in autism (Lin et al., 2012) or as in traumatic brain injury (Hsieh, 2008). SCR can also be employed when the therapeutical protocol is customized to the status of the patient, as in physical rehabilitation (Hsieh, 2008; Klein et al., 2011; Lima-Gregio et al., 2011; Ross and Elliott, 2011; Arazpour et al 2013, Peter et al., 2012; Schubert et al., 2012; Smania et al., 2012). SCR has also been considered as a possible method for the scientific evaluation of manual therapies (Wolery and Harris, 1982; Sanders, 2003) and several reports argue for this (Maricar et al., 2009; Puentedura et al., 2010; Treleaven, 2010). Recently, SCR has been used to characterize the benefits of chiropractic manipulations on neck pain and headache (Montan ˜ez-Aguilera et al., 2010; Chaibi and Tuchin, 2011). Conventional effect size cannot be calculated to estimate differences generated by a treatment vs. baseline in a single subject. Indeed, SCR avoids the use of inferential statistical tests. However, several works have proposed alternative methods for differences evaluation by using non-parametric distribution-free models (Parker and Vannest, 2009; Wolery et al., 2010; Parker et al., 2011a). Several indices can be determined from data pairwise comparison with statistical arguments, making these more convenient for estimating treatment-related changes (Parker and Vannest, 2009; Parker et al., 2011a and b). In the present work, we hypothesized that the consequences of osteopathic manipulative treatment can be evaluated through a SCR analysis. We recruited an elderly woman who complained for body instability and low back pain. The patient was treated by three consecutive osteopathic manipulative sessions. Primary and secondary outcomes consisted in monitoring the displacement of the patient’s centre of foot pressure (CoP) onto a force platform and the pain level before and after OMS.
Material and methods Patient characteristics The patient (recruited in December 2011 at IdHEO Osteopathic School, Nantes, France) was a 77 years old Caucasian woman (height Z 1.60 m, weight Z 70 kg, BMI Z 27.3). Before the recruitment, the patient had never consulted an osteopathic practitioner. The patient declared to be a housewife and reported four pregnancies in 1964, 1966, 1967 and 1970. The first one was complicated by coccyx
381
dislocation at the delivery; the last three proceeded without any complication. No perineum rehabilitation was required between pregnancy episodes. The patient did not exhibit any acute pathologies but was being pharmacologically treated for moderate arterial hypertension, vascular thrombosis, hyperlipidaemia, vitamin D deficiency and diffuse back pain episodes (over the previous six years). Treadmill walking did not reveal any cardiac dysfunctions. The patient’s husband has become hemiplegic after a cerebro-vascular thrombosis in 2005. As a consequence of this, the autonomy of the patient’s husband for feeding and personal hygiene was strongly diminished. The patient declared a strong physical involvement in the daily care of her husband. Furthermore, the patient reported two falling episodes during summer 2011. A NMR imagery of the head-neck region did not reveal any cerebral or spinal alteration. The absence of neuro-anatomical symptoms was noted. At the time of the recruitment, the patient reported low back pain in the area of both sacroiliac joints that may have been related to physical efforts necessary for the daily care of her husband. Moreover, the patient complained of a general feeling of postural imbalance, not associated with vertigo. The patient had not undergone previous postural investigations before this study. The patient was informed about the aim and the conditions of this study. She gave her informed consent according to the principles of the Helsinki declaration. The general conditions of the study were approved by IdHEO Nantes ethical committee composed by three senior osteopathic manipulators, two scientists (PhD) and one medical doctor (MD, PhD).
Monitoring of the postural control The position of the centre of foot pressure (CoP) was determined by using the WinPosturo AFP 40/16 force platform and the WinPosture NV software (Medicapteurs France SAS, Balma, France). Three strain gauges form the three tops of a virtual equilateral triangle on the platform. The patient’s feet were adjusted onto the platform by a removable plexiglass device so that the measures were realized reproducibly. In these recording conditions, the CoP position ideally corresponds to the triangle barycentre, the coordinates of which are 0 mm in both antero-posterior and medio-lateral directions. The antero-posterior and medio-lateral values of the CoP position of our patient were given as a function of this 0 mm reference. Left or right lateral shifts of the CoP position correspond to negative or positive medio-lateral values, while a front or rear shift corresponds to positive or negative values in the antero posterior direction. The patient was asked to focus on a target located on a vertical wall, at 90 cm from her eyes. Furthermore, the patient was asked not to clench her jaw and to stand as still as possible during the time of recording (Zok et al., 2008). The measures were repeated 5 times before and after each OMT session. Two consecutive measures were separated by a rest episode (duration: 30 s) during which the patient remained seated. For one measure, the CoP displacement was acquired for 51.2 s (acquisition
PREVENTION & REHABILITATION: SINGLE CASE STUDY
Can osteopathic manipulative treatment modify elderly posture
382 frequency Z 40 Hz) (Sasaki et al., 2002). This acquisition time is imposed by the manufacturer. The signal was amplified, low-pass filtered (cut off frequency Z 10 Hz), digitized (sampling frequency Z 20 Hz) through a 16 bits analogic/digital converter, and finally stored onto a personal computer for further analysis.
PREVENTION & REHABILITATION: SINGLE CASE STUDY
Outcome measurements Postural parameters From each episode of CoP monitoring, seven postural parameters were automatically calculated from the displacement of the CoP by WinPosture NV software. These corresponded to the mean medio-lateral CoP position (MLP in mm) and the mean antero-posterior CoP position (APP in mm), the surface area (S in mm2) of the CoP displacement, the total length (L in mm), the medio-lateral length (MLL in mm) and the antero-posterior length (APL in mm) of the CoP path, and the mean CoP velocity (V in mm$s1) during the CoP displacement. The displacement of CoP corresponds to body oscillations that are related to particular frequencies. The Fast Fourier transform (FFT) is a conventional mathematical procedure for the decomposition of an oscillating signal into a frequency spectrum. It allows the frequency analysis of a time-related physiological function including the displacement of CoP (Pichon et al., 2006; Cabeza-Ruiz et al., 2011). FFT calculates the energy (in V2 Hz1) associated with each frequency of the spectrum. The summation of all frequency-related energies corresponds to the power spectrum total energy (PSTE). A treatment can induce some modifications of body oscillations through modifications of the PSTE. One can analyse these modifications through the total energy (rough information) or through each frequency-related peak of energy (extremely precise information). According to previous works devoted to frequency analysis of postural control, the total power spectrum energy is distributed in three frequency bands depending on the type of somatic regulation: very low frequencies (0e0.5 Hz) correspond to the action of the sensory systems (visual and vestibular), low frequencies (0.5e2 Hz) are related to the regulation of the cerebellum, and medium frequencies (>2 Hz) reflect proprioceptive regulation (Dichgans et al., 1976; Njiokiktjien et al., 1978; Nagy et al., 2004; Bizid et al., 2009; Cabeza-Ruiz et al., 2011). In our conditions, the oscillations of the patient were automatically analysed for the three frequency bands: very low frequencies (VLF: from 0 to 0.5 Hz), low frequencies (LF: from 0.5 to 2 Hz), and medium frequencies (MF: from 2 to 5 Hz). The contribution of each frequency domain to the overall body oscillation was expressed as a percentage of PSTE. This was done for oscillations in both ML and AP directions.
Osteopathic protocol A strategy including three osteopathic manipulative sessions (OMS) was defined. The overall procedure consisted in an A-B-A-C-A-D protocol, with A corresponding to baseline measurements before OMS and B, C, and D correspond to
F. Pellerin et al. the outcome measures after first, second and third OMS, respectively. Each OMS lasted 30 min. All OMS were performed by the same French trained osteopathic practitioner (two years’ experience after an initial 5 years osteopathic training). One week separated two consecutive OMS, during which the patient was asked not to modify her way of life. First OMS (OMS01) From osteopathic tests and examination, the following points were observed : a limited mobility of the right coxofemoral joint during external and internal rotation; a restriction of the iliac bone’s posterior mobility; the coccyx tilted in the lower left direction; the right psoas muscle remained contracted and the right diaphragm dome stayed in a low position during expiration. The OMS consisted in: i) the coxofemoral and sacroiliac joints mobilization by a general osteopathic treatment of the lower limbs (the patient remained supine) (Nicholas and Nicholas, 2008); ii) the diaphragm was released by helping the diaphragm upward movements during respiration, by inserting the ulnar side of the practitioner hand beneath the patient chest, and by pushing the right diaphragm dome in a superior direction (the patient remained supine) (Nicholas and Nicholas, 2008); iii) the restoration of coccygeal mobility by applying appropriate mobilising pressure during expiration (the patient was prone) (Nicholas and Nicholas, 2008). Second OMS (OMS02) From osteopathic assessment, it was deduced there existed limited mobility of the right talus, the right external malleolus and an externally translated right patella. The right knee appeared compressed. The sacrum tilted inferiorly and the right piriformis muscle was hypertonic. The osteopathic treatment consisted in: i) a general osteopathic treatment of the right foot allowing a muscular release in the right ankle region (the patient remained supine) (Nicholas and Nicholas, 2008); ii) structural and high velocity low amplitude manipulations were performed, involving the right talus (the patient remaining in decubitus) and the right fibula (the patient lay on the left side) (Nicholas and Nicholas, 2008); iii) inhibition of the piriformis muscle was carried out (the patient remained prone) (Nicholas and Nicholas, 2008); iv) the increase of the sacrum mobility (the patient remaining in procubitus) (Nicholas and Nicholas, 2008); v) decompression of the right knee (the patient sitting on the manipulation table with the table edge inserted in his popliteal fossa) ((Nicholas and Nicholas, 2008). Third OMS (OMS03) From osteopathic tests, a lumbar restriction at the Th12-L1 and L5-S1 levels were noted, in association with a hypomobility of the sternum and the right anterior region of the chest. The right diaphragm dome stayed in a low position during exhalation, as already observed. The osteopathic treatment consisted in: i) a general osteopathic treatment of the chest (Nicholas and Nicholas, 2008); ii) mobilization of the sternum (Nicholas and Nicholas, 2008); iii) relaxation of the right diaphragm dome (Nicholas and Nicholas, 2008); iv) a bilateral general
Can osteopathic manipulative treatment modify elderly posture
Data analysis We calculated means, standard deviations (sd) and 95% confident intervals (CI95) from our data by using Sigma Plot 12.0 (Jandel Scientific). In SCR, the comparison of data sets by inferential statistics cannot be done since the independence of the values used for mean calculation is not established. Here, a data pairwise comparison was used to analyse any overlap between data sets obtained before and after OMS. We proceeded according to the method presented by Parker et al. (2011a). Actually, the results of the 5 measures obtained before and after each OMS were represented as a function of the number of the measure in a time-forward direction. Each result of the previous session was compared to each result of the subsequent session. Thus, 25 data pairs were generated for each OMS. For each data pair, the result of the subsequent session could be higher, lower, or equivalent to the result of the earlier session. Thus, the numbers of improving, declining and identical data pairs were counted for each OMS. From these, Taunovlap could be calculated as :
negative value corresponded to a decrease of the parameter value. Taunovlap value is not an estimation of the statistical significance of a difference between before and after data sets. As pointed by Parker et al. (2011a,b) Taunovlap can also be obtained from a ManneWhitney U test, from which a p value can be deduced. The different values for Taunovlap and p were calculated by using an automatic web-based calculator located at http://www.singlecaseresearch.org. We considered that p < 0.05 indicated that the difference between before and after data sets was significant.
Results Postural parameters We calculated the values of the seven postural parameters (MLP, APP, S, L, MLL, APL and V) for five CoP displacements recorded before and after each OMS. The values for means, standard deviations and CI95 are presented in Table 1. Medio-lateral and antero-posterior positions of CoP At the onset of the study, the mean values for MLP and APP were 14.0 mm and 42.6 mm, respectively (Table 1). Following OMS01, MLP and APP mean values were
ðnumber of improving pairsÞ ðnumber of declining pairsÞ Taunovlap Z100$ total number of pairs
A positive value for Taunovlap was expected when the number of improving pairs exceeded the number of declining pairs. If Taunovlap was 100%, this corresponded to no overlapping between the two data sets. Thus, a positive value indicated that the treatment has promoted a positive evolution of the measured parameter. Alternatively a
12.5 mm and 42.4 mm, respectively (Table 2). From data pairwise comparison after vs. before conditions, we calculated Taunovlap Z 44% for MLP and Taunovlap Z 28% (Fig. 1) for first OMS. Considering p values (Fig. 1), we found that these differences were not significant. Similar calculations and comparisons were made for MLP and APP values
Table 1 The values of the different postural parameters (MLP, APP, S, V, L, MLL and APL) were calculated from the CoP displacement recorded before and after each OMT session. Mean, sd, and 95% confident interval (CI95) were determined from 5 recordings obtained before and after each OMT session. OMS01
MLP (mm) APP (mm) S (mm2) V (mm$s1) L (mm) MLL (mm) APL (mm)
Mean CI95 Mean CI95 Mean CI95 Mean CI95 Mean CI95 Mean CI95 Mean CI95
(sd) (sd) (sd) (sd) (sd) (sd) (sd)
OMS02
OMS03
Before
After
Before
After
Before
After
13.4 (2.7) 16.8; 10.0 42.6 (1.2) 44.1; 41.1 127 (19) 104; 150 10.6 (1.1) 9.2; 12.0 497 (58) 425; 569 260 (36) 205; 305 383 (46) 326; 440
11.3 (2.0) 13.9; 8.7 42.4 (1.2) 43.9; 40.9 144 (14) 127; 162 9.2 (1.1) 7.8; 10.6 426 (56) 357; 495 237 (34) 195; 279 309 (48) 250; 368
10.4 (3.2) 14.3; 6.5 41.3 (4.1) 46.3; 36.3 113 (14) 95; 131 10.0 (1.0) 8.8; 11.2 470 (54) 403; 537 247 (36) 202; 292 356 (39) 307; 405
9.5 (3.6) 13.9; 5.1 38.6 (2.5) 41.6; 35.6 156 (26) 123; 189 9.3 (0.8) 8.3; 10.3 438 (45) 382; 494 237 (40) 188; 286 325 (29) 289; 361
9.5 (3.6) 13.8; 5.2 37.7 (1.5) 39.6; 35.8 120 (21) 93; 147 9.2 (0.6) 8.5; 9.9 432 (28) 397; 467 218 (24) 188; 248 330 (36) 285; 375
4.9 (1.5) 6.8; 3.0 37.5 (1.6) 39.5; 35.5 196 (41) 146; 246 8.8 (0.4) 8.3; 9.3 411 (23) 383; 439 232 (21) 206; 258 296 (13) 280; 312
PREVENTION & REHABILITATION: SINGLE CASE STUDY
osteopathic treatment in association with a stretching of the lumbo-sacral hinge (Nicholas and Nicholas, 2008).
383
384
F. Pellerin et al.
Table 2 The values of Taunovlap and p were calculated, as indicated in the methods section, by comparing data pairs obtained before and after each OMS and obtained before the first- and after the third OMS (01e03) for each postural parameter. Significant results are highlighted in bold. ns: not significant.
MLP APP S V L MLL
PREVENTION & REHABILITATION: SINGLE CASE STUDY
APL
Taunovlap p Taunovlap p Taunovlap p Taunovlap p Taunovlap p Taunovlap p Taunovlap p
OMS01
OMS02
OMS03
01 to 03
44% ns 28% ns 52% ns 60% ns 60% ns 36% ns 68% ns
12% ns 44% ns 100%
