1 Introduction

Idiopathic Scoliosis (IS) is a complex three-dimensional spinal deformity of unknown etiology and resulting in a considerable impact on morphology and movement.¹ Defined as a lateral curvature of the spine greater than 10° measured on a posteroanterior radiograph, this deformity comprises changes with lateral shift in coronal plane particularly. Two-staged process of the deformity has been discussed: a first one is noticed with a small curve and associated to defect of the neuromuscular and sensorimotor system, and a second stage dependant of the first deformity and in relation to neurological dysfunction disturb associated to recalibrate Center of Pressure position in relation to Center of Mass or CoM.² So, according biomechanical view, scoliosis influences internal mass distribution and can alter the head position. Postural control³ and dynamic equilibrium evaluated with gait analysis⁴ or freely oscillating platform in sagittal plane (personal communication) present higher Center of Mass displacements. Physiologically, a malfunction of the vestibular system, proprioceptive system and/or sensorimotor integration impairment have been noticed for patients with scoliosis,^4,5 whereas 65% of the relevant information to maintain equilibrium comes from the vestibular system.⁶ Furthermore, the maturation of proprioceptive system is consistent with the onset time of IS during the second decade, whereas an intact proprioceptive function based on the orientation of the physical vertical is essential for normal movement control and dynamic equilibrium. So, it seems interesting to develop dynamic tests which could evaluate incidence of such impairment in order to suggest rehabilitation program or brace to limit deformity. Additionally, evaluate the incidences of such malfunction may be important since the possibility of a postural disequilibrium as a contributory causative factor in adolescent Idiopathic Scoliosis has been previously suspected.^7,8

Surgery is considered necessary for cases exhibiting severe scalable deformity.¹ Surgical treatments could potentially have an effect on equilibrium⁹ and reduce movement in the segments subjected to arthrodesis.¹⁰ Spinal fusion surgery aims to correct the spinal curve, particularly in coronal and sagittal planes. In personal study, we revealed a positive impact of spinal fusion on dynamic equilibrium in sagittal plane in opposition to studies based on postural control. Schimmel et al. (2011) depicted none influence of spinal fusion on sensory system immaturity and motor-sensory integration with six quiet standing tasks.¹⁰ But this result could be in relation to experimental conditions: subjects were positioned directly on platforms that limited body oscillations, in particular head oscillations. The vestibular system has a role of controlling variations of head position at the quite standing posture. In function of spine deformity and spine stiffness (associated to spinal fusion), Center of Mass of the trunk is modified that influences variations of head position and vestibular system.⁴ So, Schimmel's results could be in relation to low differences between body oscillations before and after fusion. Specific imbalance in frontal plane with dynamic equilibrium and specific tests could reveal differences between controls and patients with IS.

As evoked, spinal fusion could be associated to spine stiffness and back stiffness.¹⁰ Anchoring Indices (AIs) are considered an indicator of postural control strategies^3,11: the AIs makes it possible to identify which reference segment is being used. In case of gait analysis, such back stiffness has been depicted for patients with IS after spinal fusion. So, this stiffness could modify sensory reference and explained variations on dynamic equilibrium.

Pre-surgery adolescents with IS present several deficiencies and impairments such as balance control and perceptual problems.^1–4,10,12 However, these results dependent of studies: patients with IS exhibited only tendencies of reduced direction control during the limit of stability tests.¹² Postural control and dynamic equilibrium assessments are obtained by measuring the sensory inputs through sensory organization tests under varying conditions: eyes open or closed, different foot positions, and frequency of oscillations.^3,5 Free-oscillating platform with significant radius and higher frequency platform movements (promoting the vestibular information) than dynamic stabilometry have been introduced in case of athletes with specific normative data.¹³ In relation to fear of falling, these free-oscillating platform revealed specific information on patients with IS before and after spinal fusion. Fear of falling may accentuate balance control strategies (essentially based on vision, proprioception or vestibular information) and reflect an accurate knowledge about the skills used to avoid falling.¹⁴ Lateral oscillations need uprightness essentially. Proprioception alone is depicted sufficient for a reasonable estimate of uprightness and ‘reliable’ vestibular input increases sensitivity.¹⁵ But in function of the accelerations of the platform (in relation with frequency of oscillations) and the vestibular canal's perception threshold,^3,16 vestibular input was principally used. So, considering these lateral oscillations, free-oscillating platform with significant radius and specific conditions could quantify the evolution of specific balance control strategies in function of spinal fusion for patients with IS. Furthermore, it seems necessary to evaluate CoM oscillations for such dynamical analysis and compute variables from CoM.¹⁷ A previously analysis revealed significant differences on displacement of CoM between patients with IS (before spinal fusion) and asymptomatic population while lateral oscillations were studied and subjects had eyes open and feet in a parallel position on a surface defined by a rectangle (as detailed on previously personal communication). Since patients with IS demonstrate a malfunction of the vestibular system, proprioceptive system and/or a sensorimotor integration impairment, one other condition of study in relation to sensory organization could be evaluated too. Patients should describe higher displacements and higher velocity of CoM, associated to back stiffness quantified with anchoring index.

So, the present study aimed to investigate the consequence of scoliosis and spinal fusion on dynamic equilibrium in frontal plane with specific condition (visual deprivation) in reference to asymptomatic subjects. With retrospective approach, the present two-phase study (before and one year after spinal fusion) tested four hypotheses: (1) displacements of CoM is significant higher in dynamic equilibrium on an unstable platform oscillating according lateral axis between a population with Idiopathic Scoliosis before spinal fusion in reference to controls. (2) In relation to malfunction of the vestibular system and sensory integration, displacements of CoM and velocity of CoM for specific vestibular condition differ in dynamic equilibrium between patients and a healthy population. (3) Anchoring Index between upper and lower back is lower in dynamic equilibrium on an unstable platform oscillating according lateral axis between a population with Idiopathic Scoliosis and a healthy population because of back stiffness. (4) Spinal fusion decreases the dynamic equilibrium differences between these populations.

2 Method

2.2

2.2 Experimental setup

A freely oscillating platform was used (Fig. 1). With this platform, oscillations were recorded in the frontal plane. Two positions were observed with platform roll: eyes open and feet in a parallel position on a surface defined by a rectangle (Condition A-evaluation of the influence of vision, proprioceptive and vestibular information); eyes closed and feet in a parallel position (feet apart) on the same surface (Condition B-evaluation of the influence of vestibular and proprioceptive information). These two conditions correspond respectively to Condition 4 and Condition 5 in our previous study published in this journal.

Close

Fig. 1 Experimental conditions (left). Platform: top and lateral view (right).

Export to PPT

These conditions were assigned at random. The experimenter held the platform (while manually obscuring a marker placed on the platform, thereby preventing recording), the subject mounted the platform and stood in the requested position. The experimenter let go of the platform, which defined the starting point of the trial. Each subject was asked to remain in equilibrium for 10 s in a prescribed position. The subject's oscillations were recorded during this time. Vertical posture was not compulsory, although foot movement was not permitted. After 10 s had elapsed, the experimenter held the platform while obscuring a marker placed on the platform, which represented the end of the trial. If the subject was unable to remain in equilibrium during the allotted time, then the actual test duration was recorded. Each subject was given 1 min to rest between tests.

3 Results

Characteristics of the IS group are shown in Table 1 in reference to Lenke's classification.²¹ Twenty four control subjects are studied (mean of age 20.7 years (SD 1.5); mean of height 166.7 cm (SD 8.2); mean of weight 59.5 kg (SD 6.6); mean of BMI 21.2 (SD 1.8)). The demographic data were similar between patients with scoliosis (before fusion) and asymptomatic population. Cobb angle values for the patients with scoliosis were significantly reduced after spinal fusion. One year after spinal fusion, patients with IS were taller but no significant difference was noticed on Body Mass Index.

3.2

3.2 Anchoring index

Fig. 2 presents the segmental AI with median and quartile values for each condition and population. After spinal fusion, patients revealed reduced variability: they tended to adopt similar mobility between upper and lower back. With Condition A, a significant group effect was noticed (H = 6.22, p = 0.04). In opposition to after spinal fusion, patients with IS presented significant higher AI than controls before spinal fusion. None significant group effect was observed with Condition B (H = 5.44, p = 0.06).

Close

Fig. 2 Anchoring Index between upper and lower back for each group and condition. “Cont”: control group; “IS bef”: patients with IS before spinal fusion; “IS aft”: patients with IS after spinal fusion. # Statistical difference between IS before spinal fusion and control population for the same condition. * Statistical difference between IS one year after spinal fusion and control population. † Statistical difference between before and one year after spinal fusion. Statistical difference with 0.

Export to PPT

The differences between AI and zero revealed significantly positive values in case of patients with IS for each condition before spinal fusion, and only for Condition A after this fusion.

4 Discussion

For patients with Idiopathic Scoliosis, the potential impact of scoliosis and spinal fusion were investigated by measuring displacements of CoM, velocity of CoM and Anchoring Index between upper and lower back during dynamic equilibrium with lateral oscillations in reference to asymptomatic subjects. Using a free-oscillating platform presenting significant radius, the potential impact of visual deprivation was specifically used. Considering eyes opened, with a positive and limited impact of spinal fusion on displacements of CoM, this result is associated to a specific organization between upper and lower back (without back stiffness). Spinal fusion tended to reduce displacements and velocity of CoM. Considering that eyes closed may accentuate vestibular control (whereas patients with IS present a deficiency of the vestibular system), and based on CoM measurements, results reveal none significant difference between control group and patients with IS. In case of such platform with important radius, dynamic equilibrium would essentially be based on vestibular information, ‘reliable’ vision and proprioceptive input increases sensitivity. So, with lateral oscillations, these results confirm and precise Bisdorff's results¹⁵: in case of patients with IS, a malfunction of the sensory integration was observed. The difference noticed on displacements and velocity of CoM between patients and controls could be associated to lower sensory integration, that limit impact of ‘reliable’ vision and proprioceptive information on equilibrium. The difference between results on Condition A and Condition B (in reference to asymptomatic subjects) would suggest that a sensory conflict (related in particular to vision) would disrupt the control of the CoM. Moreover, patients with IS use vertical axis as reference for lower back before spinal fusion certainly in relation to proprioceptive information. This surgery modifies such reference (without back stiffness) and promotes better dynamic control.

We considered dynamic equilibrium as a precarious balance condition: somatosensory cues allow an acute perception of spatial orientation.¹⁶ So, in contrast to Schimmel et al. (2011) who display a condition of postural equilibrium, patients with IS presented poorer dynamic equilibrium compared to healthy controls.¹⁰ With eyes closed, lateral oscillations revealed no significant difference for displacement and velocity of CoM between healthy population and patients with IS. Patients and asymptomatic subjects maintained good dynamic control through a greater contribution of another sensorial input (vestibular, proprioception or somatosensory system), even if information on the relative position of the body with respect to the external environment was considered as blocked.^1,22 This result is phase with previous results³ in case of very slow oscillations of the support. Assaiante et al. (2012) used a motorized rotating platform. In their approach, the frequency applied let to control specifically sensory information used to maintain posture. In their results, Assaiante's et al. revealed that proprioceptive information in sensory integration of postural control can be neglected in case of patients with IS. Our results revealed that other sensory information can disturb sensory integration too. With a moveable platform, Guillou et al. (2006) revealed that recovering balance with such lateral oscillations is rather a motor component than an equilibrium one.¹⁶ In our study with dynamic platform, it seems difficult to define motor component to equilibrium one, even if use vertical axis as reference on proprioceptive information let to limit mobility of the trunk or head in order to control displacements and velocity of CoM. About the equilibrium component, in case of rocking platform and asymptomatic subjects (with no previous history of orthopedic, vestibular or neuromuscular disease), Pozzo et al. (1995) described that head or trunk can be used as reference in function of mechanical and sensorimotor specificities, of motor abilities or of tasks.²³ So, specific coordination between upper and lower back in case of patients with IS before spinal fusion could constitute a part of this reference. About the motor component, when considering this specific task of lateral oscillations, it is evident that even simple movements require complex control. Upper and lower back coordination could be one of mechanisms for simplifying the complex control of equilibrium with specific synergies.

Using the same low-cost platform and other variables, Mesure et al. (1995) compared dynamic equilibrium between six athletes and nine novices.¹³ In opposition to this study, vision privation is not based on darkness condition. Such condition is associated with an increased risk of falling. This condition let to evaluate the influence of vestibular information with low-sensitivity: feet in parallel position limit displacement of trunk. Feet in “tandem” position (heel-to-toe foot position) with eyes opened/closed could certainly be more sensitive conditions as it was noticed in case of anterior-posterior oscillations (personal condition in this journal). Nevertheless, this particular condition is very unstable and difficult in case of lateral oscillations because axis of feet is parallel to the pivot of the platform.

Based on velocity of CoM for Condition A, the sensory system in IS patients might still have been immature one year after surgery: information processed by the central nervous system was only partially exploited and no significant differences were noticed between patients after spinal fusion and controls. Spinal fusion influenced partially CoM displacements in case of postural control.¹⁰ In accordance with these authors, spinal fusion might not compensate for sensory system immaturity and motor-sensory integration in case of dynamic oscillations. We could hypothesize that none specific sensory system maturity or none new motor-sensory integration in relation to physiologic and natural evolution was appeared after spinal fusion (patients were captured twelve months after spinal fusion). Several compensatory mechanisms might be implicated depending on the kind of input received and how this information is then processed by the central nervous system.¹⁰

Power calculation and number of subjects were performed for describing differences from displacement of CoM for asymptomatic population and patients with scoliosis before spinal fusion in Condition A. In relation to preliminary personal communication and in reference to preliminary article based on anteroposterior oscillations, this condition has been estimated as sufficient to describe significant difference between asymptomatic controls and patients before spinal fusion. Furthermore, power calculation and number of subjects are based on displacement of CoM considered as an indicator of postural performance. This performance quantifies the ability to ensure dynamic stability in challenging conditions.

Defined as a point around which the object's mass is equally distributed in all directions, CoM has been currently used in literature.^1,4,7,10,13 From Center of Mass, mean displacement and mean velocity quantified dynamic equilibrium according: the smaller the value, the better the dynamic control. Our results reveal a partial positive impact of spinal fusion with Condition A. Even if dynamic oscillations are more discriminating than static condition,^10,16,23 mean velocity was not significantly reduced in Condition A after surgery. Spinal fusion did not reduce significantly displacement of CoM likewise, but tended to lower values. So, in case of dynamic equilibrium with lateral oscillations, displacement of CoM seems to be an indicator of postural performance.

In case of static and dynamic analysis, Quotient's Romberg could be analyzed.¹³ None Quotient's Romberg was calculated in this article. Such Quotients are calculated as the Ratio between values calculated from data obtained with two conditions (Closed and Open Eyes values for instance). However, using ratios need a particular attention: the type of relationship between the numerator and the denominator, the potential intersection with the origin influence these ratios.²⁴ Wrong clinical interpretations are possible. Therefore, these seductively simple ratios could be analyzed in a specific study.

Nevertheless, the present study had the following limitations. First, about the Condition B, the experimenter asked subjects to close their eyes. Contrary to other methods of dynamic equilibrium analysis,¹ blindfolds were not used during our study, and subjects were not maintained in position with belts, which could have increased confidence. This point could explain the results obtained on Condition B.

Second, our approach used a particular condition of equilibrium: a freely oscillating platform with a large radius and variables were chosen in accordance with classic postural analysis. Considering that dynamic equilibrium can be quantified with Center of Mass, Center of Pressure or anchoring index applied on segments,³ this retrospective study is only based on parameters computed from CoM and anchoring index. Moreover, these parameters could be associated to mobility of several segments like pelvic, upper back, middle back, lower back and head computed from kinematic data in order to define strategies.

Third, this pilot study is based on a limited number of subjects, the use of a healthy population as control presents a first normative approach with durations, anchoring index; displacements and velocities of CoM. About patients with scoliosis, only spinal fusion has been studied. It could be interesting to evaluate incidence of spinal brace.

5 Conclusion

In case of lateral oscillations, spinal fusion is associated to a partial better dynamic equilibrium in relation to better sensory integration one year after surgery. We evoked postural disequilibrium as a contributory causative factor in adolescent idiopathic scoliosis⁷: so, our results could imply an increasing improvement after this first year. A new study to evaluate the effects of surgery at 10 years of follow-up could be very interesting.

Funding

None.

Ethical approval

Ethics Committee of Angers (n° 2017/08).