Functional impairments are common among LSS patients [
3,
6,
9]. Some authors have suggested [
4,
30] that these impairments may be related to peripheral sensory loss. However, it remains unclear which postural mechanisms are directly involved. Possible candidates are impaired motor actuator, sensorimotor time delay, impaired weighting of visual, vestibular and proprioceptive input, passive muscle elasticity, and damping. This ambiguity motivated us to conduct the present study. Our study’s aim was to identify the postural control parameters that may reveal underlying sensorimotor mechanisms responsible for LSS patients’ balance problems. Moreover, we aimed to monitor pre-post-surgical changes in these parameters. We hypothesized that the main effect in LSS patients concerns the altered weighting of sensory information. In particular, we expected an altered use of proprioceptive information in LSS patients as compared to healthy control subjects.
Since this study serve as basis for further investigations aiming to understand the underlying sensorimotor mechanisms responsible for LLS patients’ postural instability, sample size was not based on a previous power calculation. The small sample size may limit our study’s results regarding stronger differences between groups. Furthermore, we propose to monitor post-surgery physiotherapy measures or to intervene specifically in order to control influencing factors on postural stability, and to expand the observation duration to detect long-term effects.
Spontaneous sway
Spontaneous sway measurements revealed that LSS patients’ postural sway was greater, faster and at higher frequencies compared to matched healthy control subjects. Those alterations have been also described in LBP patients [
18‐
29] and elsewhere in LSS patients [
4,
5]. Abnormal spontaneous sway was associated with a higher risk of falling in LSS patients [
7]. Our PROs confirmed functional limitations in LSS that considerably affect patients’ quality of life, and additionally indicate stress-dependent worsening of impairments.
Post-surgery, postural sway (as measured by RMS) was reduced and approximated the values of control subjects which paralleled findings from an earlier study [
15]. This reduction may indicate that patients achieved better postural stability post-surgery following the decompression of lumbar radices, which carry the motor and sensory information for lower extremity control. Specifically, the RMS reduction after surgery and rehabilitation measures may be a consequence of improved peripheral proprioception. Since surgery significantly reduced pain intensity and patient-reported functional limitations, as expected, this also may contribute to improved stance stability, which is also influenced by rehabilitation measures.
Perturbed stance and model-based analysis
In addition to investigating spontaneous sway, we assessed postural reactions to external stimuli. Specifically, we compared patients’ postural sway to continuous pseudorandom platform tilts in anterior-posterior direction. These raw angular traces were transformed into transfer functions in the frequency domain, reflecting patients’ reactive body excursions as a function of stance perturbation. These transfer functions revealed a smaller GAIN in patients than control subjects, especially in lower and mid-range frequencies. In other words, patients’ postural reactions were, on average, smaller than control subjects’. Furthermore, patients presented a smaller PHASE lag, representing a relatively earlier postural reaction to external stimuli than control subjects. Interestingly, PHASE differences between patients and control subjects were amplitude-dependent (apparent at 0.5°, but vanished at 1°). Patients’ PHASE lag at 0.5° corresponded to that of control subjects at 1°, meaning that patients reacted to smaller platform excursions like control subjects to greater excursions. Post-surgery, patients approximated the PHASE behavior of control subjects, which is mainly based on changes at 0.5 rotational amplitude and at hip movement.
On a first glance, LSS patients’ smaller and earlier reactive body excursions, as compared to control subjects, do not seem to be disadvantageous for postural stability. Even more surprising, post-surgery we observed a prolongation of postural reactions in LSS patients, represented by an increased PHASE lag. How could this prolongation be related to the fact that patients achieved enhanced posture stability post-surgery, as our spontaneous sway findings demonstrate (RMS)?
The transfer function findings reported above formed a basis for a model-based data interpretation: Extracting model-parameters after fitting simulated to experimental transfer functions revealed slightly more pronounced velocity control (damping) in the postural system (Kd) in patients compared to control subjects. Kd results are in line with the patients’ aforementioned PHASE behavior. Furthermore, patients use significantly less proprioceptive information, as reflected by smaller values for proprioceptive weight (Wp), than control subjects. No other model parameters differed between groups.
Following the assumption that LSS patients weigh their sensory information differently than healthy subjects, we hypothesized that patients rely relatively more on space cues and less on proprioceptive cues, as they suffer from stenosis-induced afferent dysfunction. Our GAIN results support this assumption, since our patients revealed a smaller GAIN, indicating their stronger orientation towards space coordinates and less dependency on platform movements [
31,
43]. In contrast, healthy subjects integrated more proprioceptive information for stance regulation resulting in a greater GAIN. These findings are emphasized by our model-based approach, where we found patients’ proprioceptive cues (Wp) down-weighted and as a result, space cues implicitly up-weighted when compared to healthy subjects. However, up-weighting space cues implies a functional disadvantage, as the vestibular system carries a larger amount of sensory noise [
46], leading to less accurate stance regulation. There is evidence that larger platform perturbations generally lead to a relative smaller GAIN, stronger orientation on space cues, a smaller PHASE lag, and greater velocity control (Kd) [
47] to ensure stance stability. We assume that patients’ postural behavior might represent a compensation strategy, where patients try to prevent excessive and rapid changes in postural position already at small excursions by reducing PHASE lag due to their stenosis-caused physiological limitations and related pain. Additionally, as mentioned earlier, afferent information might be inadequate, which could be a reason why patients down-weigh proprioception. Patients compensate for their afferent dysfunction by staying within a more stable and secure range of regulation [
48], while accepting disadvantages in the accuracy of postural control through increased vestibular noise. Surgery alleviated physiological limitations responsible for this compensation, leading to a PHASE-shift towards healthy subjects’ postural behavior.
The concentration of the intervention-induced PHASE effect on the hip joint may indicate, in addition, a change in hip-strategy [
49,
50]. Concerning LBP patients, some authors reported that these patients exhibit a less effective hip strategy than healthy subjects [
24,
27,
51]. According to those findings, we speculate that patients stiffened their hip joint pre-surgery by enhancing muscular co-contractions due to the aforementioned worries about safety. Post-surgery, this stiffness may be relaxed to enable larger body excursions [
50], as surgery directly alleviates their primary impairments, eg pain [
3,
9,
52].
Considering patients’ afferent dysfunction before surgery, one would assume that motor latency as reaction of the perturbation would be prolonged. However, time delay (Td) extracted from the model-based approach did not differ between groups, highlighting the assumption that altered PHASE behavior as a function of time is not primarily caused by physiological impairments but rather by a compensation strategy, ie altered timing for stance regulation.
Additionally, the model-based total effort to correct the difference between actual and desired body angle (represented by the model parameter Kp) was similar in both groups, indicating that the intensity of motor response was also not unaltered in LSS patients. We therefore assume that the stenosis-caused efferent dysfunction was less pronounced in our patients and did not lead to a fundamental deficit in relation to their stance regulation.
Correlations
When comparing our experimental results with clinical tests of motor function and patient-reported outcomes, one is tempted to associate the spontaneous sway abnormalities, ie larger (RMS) and faster (MV) postural sway to the other functional disadvantages pre-surgery. In the literature, these spontaneous sway abnormalities are reported to be associated with postural instability and functional impairments [
4,
53]. Interestingly, our spontaneous sway parameters neither correlated with patient-reported outcomes, nor with functional tests, nor with measures derived from perturbed stance. In contrast, measures from perturbed stance significantly correlate with both patient-reported functional outcomes (FFb-H-R) and specific functional tests, which in turn correlated strongly. The significant correlation between FFb-H-R and many functional tests (PC-FT: POMA [
34], BBS [
35], CTSIB [
36], TUG-, CRT- and 10 m walking test [
38]) indicate that these functional tests mainly determine the perception of functional impairments in patients [
54], while other tests, ie the Functional Reach test [
37], or spine mobility (distance between the fingertips and ground) seem to reflect other functionality dimensions. We claim that the close correlation between our experimental results from perturbed stance, many functional tests (PC-FT), and the patient-reported deficits may be applied in future to lower the number of tests necessary to accurately assess all facets of postural impairments.
In general, our data lead us to conclude that patients’ postural reactions following external perturbations are suitable to precisely quantify most of their postural impairments unlike spontaneous sway measures. However, taking spontaneous sway measurements is the current standard when assessing patients’ functional status [
18‐
25,
29]. Based on our results, we suspect that parameters recorded during quiet stance fail to capture patients’ gait and stance disturbances sufficiently - thus we strongly propose applying external perturbation when assessing patients’ balance ability.