Fear-avoidance beliefs, anxiety and depression are associated with motor control and dynamics parameters in patients with chronic low back pain

Abstract

Purpose

To evaluate the influence of fear-avoidance beliefs, anxiety and depression on dynamic and motor control parameters before and after Functional Restoration Program.

Methods

Patients were divided into three groups depending on clinical scores scores. Dynamic and motor control parameters were extracted from gait analysis.

Results

Patients showed different ways based on preferred walking speed, capacity benefit, peaks of propulsion and gait stability in function of clinical scores.

Conclusion

Fear-avoidance beliefs associated to anxiety and depression influence biomechanics on overground walking. We depicted different ways to adapt their gait in function of biopsychosocial scores. Functional Restoration Program influence these ways.

Highlights

• •An indicator of walking capacity for chronic low back pain.
• •Influence of functional program according to subgroups.
• •Low FABQ score and gait stability.
• •High FABQ, low HAD scores and capacity benefit.
• •High FABQ, HAD scores and resilience.

1. Introduction

Chronic Low Back Pain (cLBP) remains a prevalent and persistent problem in our societies because this disability involves inability to work.¹ Subgrouping patients has been suggested as a method to implement targeted treatment approaches and investigations.² It seems necessary to understand the various impacts of this symptom and their relations in order to adapt treatment to each patient.³ Patients with cLBP show behavioral reactions like fear-avoidance, physiological reactions (muscle hypertonia), and cognitive reactions (catastrophic ideation)2, 3, 4, 5 clinically evaluated with scales. Recently, it has been suggested that patients with cLBP could describe specific motor control alteration depending on Fear-Avoidance.^6,7 Veeger et al. described this motor control alteration on walking with imposed speeds (using treadmill) without treatment applied.⁷

Even if the reasons for cLBP are still under discussion, Functional Restoration Programs (FRP) have been suggested for treating this symptom.³ FRP include selective exercises involving muscle strengthening, muscle stretching, control of trunk muscles, and aerobic exercises.³ These programs tend to emphasize “well behaviors” and prevent “sick behaviors”, without specific motor control exercises.⁴ A biopsychosocial approach is encouraged and focused on improvement in accordance with the profession and the return to work.⁴ Concerning fear-avoidance, the Fear-Avoidance Model or FAM is a cognitive-behavioral report that explains why a minority of acute low back pain sufferers develop a chronic pain problem and why this chronicity continues.⁵ As a result, patients with cLBP who describe important fear of movement need to base their rehabilitation on trust in order to reduce fear-avoidance or fear-escape that tends to improve pain experience.⁵ With a specific and multimodal approach, FRP tends to reduce pain-related fear that negatively influences pain experience and limits movements. Biopsychosocial scales are used to depict pain intensity (with Visual Analog Scale or VAS), to evaluate fear of pain (with Fear Avoidance Beliefs Questionnaire or FABQ), and to detail anxiety and depressive disorders (with Hospital and Anxiety Depression scale or HAD).⁸

In addition, as reminded by Pakzad et al.,² it seems important to establish relations between these biopsychosocial scales and the real impact of cLBP on life. Thus, biopsychosocial scores could directly be related to locomotor compensations. Yet, with treadmill walking, patients with cLBP revealed no influences of fear-avoidance beliefs on trunk movement variability during gait such as described by Veeger et al.⁶ Even if mean values for spatio-temporal parameters and peaks of trunk velocities were statistically equivalent between overground and treadmill walking, gait variability was reduced on the treadmill walking with asymptomatic subjects.⁹ Therefore, Veeger et al.‘s conclusions could be in relation to treadmill walking. Furthermore, Veeger et al. limited their approach to patients with high and low FABQ score. However, in their model, Vlaeyen and Linton (2000) described states of disuse, depression and disability as possible consequences of pain-related fear, avoidance and hypervigilance.¹⁰ Anxiety and depression could be considered as a worsening of the patient's biopsychosocial state. Recently, Weiner et al. demonstrated significant associations between anxiety or depression and gait speed in older adults with cLBP.¹¹ Then, as fear-avoidance beliefs, anxiety and depression could influence motor control in overground walking in the case of patients with high Fear-Avoidance Beliefs and intensify motor control disorders.

So, walking is described as one of the abilities most affected by cLBP.^12,13 Preferred walking speed is a spatio-temporal parameter that differentiate the gait between patients with cLBP and healthy controls, is in relation with disability, presents benefit after FRP.^3,12,14 So, difference between preferred walking speed before and after FRP can be defined as capacity benefit. Furthermore, dynamic gait parameters and motor control are associated: patients with cLBP walk with a specific motor control according to a strategy to protect themselves against the onset of pain11. Specifically, pain distribution influences the vertical Ground Reaction Force (GRF) during walking.¹² Classically, from GRF, several peaks of curves can be related to motor control.¹³ Thus, these peaks give an important information on the dynamic of movement and motor control applied and could certainly help to understand how patients with cLBP and severe or moderate disability had different ways to normalize their gait. Lateral and vertical peaks have been particularly used in the case of patients with severe stroke results in primary motor control problems for instance Ref. 15. Adequate motor control allows low variability and good locomotor stability.¹⁶ No studies have yet investigated gait stability with such population on overground walking considering fear-avoidance, anxiety and depression. ‘‘Stable gait’’ may refer to the repeatability of walking.¹⁶

In this pilot study, the objective is to demonstrate the influence of Fear-Avoidance Beliefs, anxiety and depression on dynamic parameters and gait stability for patients with cLBP in overground walking before and after FRP. The hypotheses formulated are:

• •A significant group effect is noticed on the preferred walking speed before and after FRP.
• •A significant group effect is described on the capacity benefit.
• •A significant group effect is observed on peaks extracted from GRF and gait stability of Center of Mass before and after FRP.

2. Material and method

2.1. Population

This retrospective and descriptive study concerns patients over 18 years old having cLBP, who underwent a five-week rehabilitation program (based on Mc Kenzie method according self-treatment through posture correction and repeated exercise movements at end-range performed with high frequency¹⁷). Patients were included in the study from 2013 to 2018 if they had cLBP (over 3 months). This retrospective study has been approved by the Ethical Committee of Angers (no. 2021–064). All participants signed informed consent forms to use anonymous data.

Fifteen patients with high FABQ score and high HAD score (HHAD group: FABQ≥33 and HAD>16), fifteen patients with high FABQ score and low HAD score (LFAD group: FABQ≥33 and HAD<16), and fifteen patients with low FABQ score (LFABQ group: FABQ<33) constituted our population. Regarding FABQ score, the cut-off value of 33/66 was chosen for two reasons: firstly, the mean normative data in the case of cLBP is 32.6.¹⁸ Secondly, based on the geometric average value, we computed similar normative data from previous data. FABQ scale is based on 16 items divided into two subscales (the Physical Activity subscale or FABQ-pa, and the Work subscale or FABQ-w). Global FABQ score is clinically interpreted as follows: higher scores reflect greater levels of fear-avoidance.¹⁹ Concerning HAD scale, Bjelland et al. (2002), through a literature review of a large number of studies, identified a cut-off point of 16/42 for HAD.²⁰ HAD scale evaluates the dimensions of anxiety and depression from 14 items.²⁰

For all patients, the pains described are located between the lumbar region and the lumbosacral region. No patients had any root symptoms. The rehabilitation program included physiotherapy, occupational therapy, psychology, and adapted physical activity (five days per week, during five weeks). The inclusion criteria were: patients with cLBP, without neurological or orthopaedic disorders, no surgery that would interfere with locomotion, and completed the program in its entirety. Each patient present data set was exploitable. Patients were excluded if they reported with mental retardation, musculoskeletal/neurological diseases, data set not exploitable, inequalities in leg length, locomotion injuries, surgery of the spine in the previous 3 months, or leg pain.

2.2. Experimental protocol

During the first and last week (five weeks later) of FRP, patients were asked to complete the pain and disability self-report questionnaires. Clinical gait analysis was next performed using a motion capture system (VICON, Oxford Metrics, Oxford, UK, 100 Hz) and 31 retro reflective spherical markers (14 mm) were placed by a biomedical engineer with previous experience in this type of analysis. Each equipment was used as part of the “Plug-in Gait Full Body” model. The software Nexus (version 1.8.5, Vicon, Oxford, UK) captured markers, defined gait events, and applied the model. These markers were directly placed onto the skin on: bilateral anterior superior iliac spine; sacrum, considered as the middle of the segment between posterior superior iliac spines; lateral tibial plateau; lateral malleoli; second metatarsal head; heel; lateral epicondyle of the femur; bilateral tibial wands; left and right femoral wands; bilateral shoulders (acromio-clavicular joint); bilateral elbows; bilateral wrists (radius-styloid process and ulna-styloid process); each hand (dorsum of the hand just below medial head of metacarpal); and spinous process of the 7th cervical vertebra and 10th thoracic vertebra. Two force platforms (AMTI, MCA model, 1000 Hz) were used to record Ground Reaction Forces (GRF). The force platforms were located at midpoint of the walkway to ensure constant velocity was achieved.

After the placement of the markers, patients walked at spontaneous speed (preferred walking speed) in a walking corridor defined on the ground by two lines. No instructions were given to patients on how to properly place their feet on each force platform: the patients were instructed to direct their attention straight ahead. The starting position was adjusted until the patient struck the force plate during a normal stride. A stride was considered only if the support on the platform was complete. In case of double support on platform, the stride was not retained. For each patient and therapeutic phase, the average value of five strides was considered representative and used in the subsequent analysis.²¹

2.3. Measured parameters

2.3.1. Pain evaluation and biopsychosocial parameters

Aside from the FABQ and HAD scales, patients completed the Visual Analog Scale (VAS) to evaluate Pain Intensity as follows: 0–1 = no pain, 2–3 = mild pain, 4–5 = moderate pain, 6–7 = intense pain, and 8–10 = extreme pain.

2.3.2. Biomechanical parameters

Preferred walking speed (expressed in m s⁻¹), corresponding to the average speed over the locomotive cycle, was calculated. Capacity benefit, defined as the difference of this preferred walking speed before and after FRP, was computed too to evaluate gain of FRP.

From the GRF (expressed as a percentage of the body weight), four peaks were calculated in relation to motor control: maxima on lateral GRF during the first part of the stance phase (between 0% and 50% of stance phase-noticed PX1), maxima on lateral GRF during the second part of the stance phase (51% of stance phase to toe off-noticed PX2), two maxima on vertical GRF (respectively between 0% and 30% of stance phase-noticed PZ1-and between 31% of stance phase to toe off -noticed PZ2). These four dynamic parameters have been calculated using Matlab-routine (version R2012, MathWorks Inc., Natick, MA, USA).

Gait stability was obtained by computing the root-mean-square (RMS) accelerations on the Center of Mass, calculated from the Plug-in-gait model. These data were analyzed after low-pass filtered them at 20 Hz with a fourth-order Butterworth filter. The RMS value of each of the three components of acceleration (antero-posterior, lateral, vertical) were computed as follows:

$$RMS=\sqrt{\frac{{\sum }_{i=1}^N{a}_i}{N}}$$ (1)

where RMS is defined per axis, a_i is the acceleration measured at the i-th sampled value. We considered normalized RMS: we divided RMS by the square of the walking speed and multiplied it by the average step length. This normalized RMS is dimensionless.¹⁶ RMS values increased as walking ability decreased: high acceleration values are associated with decreased gait stability.¹⁶

2.4. Statistical analysis

The results were presented as follows: mean (standard deviation). The analyses were conducted with Statistica (V13, Dell, USA; alpha level of significance = 0.05). A Shapiro-Wilk test ensured normality for each of the distributions. As most of the parameters were not normally distributed, the following non-parametric tests were performed:

• •Kruskal–Wallis H-tests on all estimated parameters. When a significant “group” effect was found, pairwise comparisons were performed using the independent Mann–Whitney U tests with the Holm-Bonferroni correction.
• •Ratios (gender) were compared with Khi².

3. Results

3.1. Clinical comparison

Forty-five patients constituted our population. Sixty one percent of the forty-five studied patients were female. The mean height and mean weight were respectively 1.69 m (0.09) and 69.41 kg (13.41) before FRP. The mean height and mean weight were respectively 1.686 m (0.09) and 69.06 kg (12.68) after FRP. The mean age (before FRP) was 38.25 years old (8.71). Regarding subgroups, no significant differences on age (p = .49), height (p = .15) and weight (p = .30) were noticed. The results revealed: LFABQ group before FRP (mean age: 39.67 years old (9.25); mean weight: 71.97 kg (13.07); mean height: 1.715 m (0.09); 46.7% were female); LHAD group before FRP (mean age: 37.93 years old (8.95); mean weight: 65.70 kg (13.04); mean height: 1.653 m (0.08); 80.0% were female); HHAD group before FRP (mean age: 36.44 years old (8.73); mean weight: 71.33 kg (14.68); mean height: 1.703 m (0.10); 55.5% were female). The LHAD group had a significant higher proportion of women (p = .0001 compared to the HHAD group and LFABQ group). About this proportion, no differences were observed between the HHAD and LFABQ groups (p = .23).

Table 1 depicts clinical scales per group, before and after FRP. Depending on the group, the clinical gains of the FRP on the FABQ, HAD and VAS scores varied. Only the HHAD group showed significant differences on these three clinical scores after FRP. Concerning HAD score or FABQ score, the LFABQ and LHAD groups revealed significant differences compared to the score evaluated before FRP.

Table 1.

Descriptive statistics of the self-report questionnaires before and after FRP for each group. LFABQ: group with low FABQ; LHAD: group with high FABQ and low HAD; HHAD: group with high FABQ and high HAD. *: significant difference at p ≤ .05.

Scale	Group	Before	After	p-value
FABQ	LFABQ	21.00 (7.96)	27.25 (16.41)	.14
	LHAD	45.83 (9.87)*	34.09 (18.03)*	.03
	HHAD	47.50 (9.35)*	38.23 (13.42)*	.02
HAD	LFABQ	16.20 (6.52)*	10.87 (6.18)*	<.001
	LHAD	13.00 (3.79)	10.91 (5.01)*	.08
	HHAD	25.77 (4.22)	18.84 (8.73)	<.01
VAS	LFABQ	5.23 (2.32)	3.75 (2.94)	.06
	LHAD	3.58 (1.59)	2.33 (1.94)	.16
	HHAD	5.85 (1.98)*	3.67 (2.35)*	<.01

3.2. Influence of fear-avoidance, anxiety and depression on preferred walking speed and benefit capacity

Preferred walking speeds and benefit capacities are detailed in Table 2. No significant group effects were observed on gait speed before FRP (H = 1.35; p = .51). A group effect was noticed after FRP (H = 9.96; p = .006). The LHAD group showed a significant higher self-selected speed than the LFABQ group (Z = −2.85; p = .004). No other significant differences were noted (Z = −1.66; p = .09 between LFBAQ and HHAD groups; Z = 1.93; p = .06 between LHAD and HHAD groups).

Table 2.

Self-selected speed and capacity benefit (defined as the difference of self-selected speed before and after Functional Restoration Program or FRP) per group. LFABQ: patients with low FABQ score; HFABQ: patients with high FABQ score; LHAD: patients with low HAD score; HHAD: patients with high HAD score *: significant difference at p ≤ .05.

	LFABQ group	LHAD group	HHAD group
Speed before FRP (m.s−1)	1.09 (0.11)	1.01 (0.20)	1.02 (0.12)
Speed after FRP (m.s−1)	1.07 (0.17)*	1.25 (0.08)*	1.16 (0.13)
Gain (m.s−1)	−0.11 (0.22)*	0.23 (0.16)*	0.12 (0.20)

Regarding the effect of FRP on preferred walking speed per group, a significant group effect is observed on this gain (H = 13.67; p = .001). Post-hoc tests revealed significant differences between the HHAD and the LFABQ groups (p = .02), and between the LHAD and the LFABQ groups (p = .001). No significant differences were noted between the HHAD and the LHAD groups (p = .99).

3.3. Influence of fear-avoidance, anxiety and depression on ground reaction forces

A significant group effect was described on peak of propulsion PZ2 before Functional Restoration Program (Table 3). The LHAD group showed significantly higher PZ2 than the HHAD group (right side: Z = −1.98; p = .04; left side: Z = −2.04; p = .04) and the LFABQ group (right side: Z = −2.41; p = .01; left side: Z = −3.39; p = .001). No significant differences were noticed between the HHAD group and the LFABQ group (right side: Z = −0.96; p = .33; left side: Z = −0.82; p = .40). After FRP, no significant group effects were observed.

Table 3.

Influence of FABQ and HAD scores (LFABQ: patients with low FABQ score; HFABQ: patients with high FABQ score; LHAD: patients with low HAD score; HHAD: patients with high HAD score) on lateral (PX) and vertical GRF peaks (PZ) for each side before and after Functional Restoration Program (FRP). GRF are expressed as a percentage of the body weight.*: significant difference at p ≤ .05.

		Before FRP					After FRP
		LFABQ group	LHAD group	HHAD group	H	p	LFABQ group	LHAD group	HHAD group	H	p
Left	PX1(N)	5.51 (1.25)	5.08 (0.98)	4.98 (1.65)	1.79	.41	5.78 (1.45)	6.41 (1.42)	6.05 (1.90)	0.55	.76
	PX2(N)	6.09 (2.09)	5.55 (1.79)	5.90 (1.63)	0.37	.83	6.02 (2.10)	5.51 (1.24)	6.48 (2.13)	0.43	.81
	PZ1 (N)	103.03 (5.94)	108.34 (13.31)	104.29 (4.35)	2.04	.36	108.88 (5.66)	108.88 (6.18)	109.38 (8.32)	0.09	.95
	PZ2(N)	103.66 (6.54)	114.49 (10.33)	106.73 (5.39)	11.15	.03*	103.92 (6.85)	107.23 (9.99)	109.19 (4.64)	5.63	.06
Right	PX1(N)	5.68 (1.37)	5.39 (1.30)	5.07 (1.55)	1.07	.59	6.06 (1.14)	5.30 (0.92)	5.61 (1.85)	2.15	.34
	PX2 (N)	6.03 (1.79)	6.13 (2.97)	5.61 (1.76)	3.75	.83	6.24 (2.00)	5.64 (1.72)	6.18 (1.95)	0.43	.81
	PZ1 (N)	103.64 (5.68)	107.79 (10.01)	103.54 (5.44)	0.66	.72	109.86 (4.97)	106.31 (5.57)	108.98 (8.68)	1.70	.43
	PZ2(N)	104.30 (6.66)	113.63 (10.98)	106.49 (4.63)	7.24	.03*	104.87 (6.29)	107.81 (8.14)	108.53 (6.32)	2.93	.23

3.4. Influence of fear-avoidance, anxiety and depression on gait stability computed on CoM

Table 4 represents the stability of the CoM per axis, before and after FRP. Before FRP, only stability of the CoM on vertical axis presented a significant difference. LFABQ group described a significant lower stability on CoM according this axis (between LFABQ group and LHAD group: Z = 0.69; p = .49; between LFABQ group and HHAD group: Z = 2.43; p = .01; LHAD group and HHAD group: Z = 0.92; p = .37). None group effects were noticed after FRP.

Table 4.

Influence of FABQ and HAD scores (LFABQ: patients with low FABQ; LHAD: low HAD score; HHAD: high HAD score) on gait stability (normalized RMS) per axes before and after Functional Restoration Program (FRP). This normalized RMS is dimensionless. *: significant difference at p ≤ .05.

	Before FRP					After FRP
	LFABQ group	LHAD group	HHAD group	H	p	LFABQ group	LHAD group	HHAD group	H	p
Antero-posterior	0.031 (0.011)	0.033 (0.012)	0.031 (0.008)	0.14	.93	0.039 (0.015)	0.038 (0.031)	0.035 (0.007)	2.02	.36
Lateral	0.035 (0.007)	0.033 (0.008)	0.031 (0.006)	1.80	.41	0.037 (0.009)	0.033 (0.008)	0.033 (0.005)	3.67	.16
Vertical	0.063 (0.007)	0.058 (0.012)	0.055 (0.008)	8.41	.03	0.067 (0.013)	0.063 (0.010)	0.056 (0.010)	4.70	.10

4. Discussion

In the case of patients with cLBP before and after Functional Restoration Program, the aim of this study was to evaluate the influence of fear-avoidance, anxiety and depression on dynamic parameters (self-selected speed, capacity benefit, Ground Reaction Forces) and motor control parameter (gait stability) in overground walking conditions. Patients with higher FABQ and lower HAD scores were significantly more sensitive to the influence of FRP on self-selected speed compared to the two other groups: their capacity benefit is significantly higher. Furthermore, patients with high FABQ score and low HAD score revealed higher propulsion peak on vertical ground reaction forces before FRP. After FRP, none group effect was described. Therefore, with specific and clinical supervision, FRP tend to limit fear-avoidance and consequences of anxiety/depression on movement. Regarding motor control approach, this result reveals two strategies according groups: patients with lower FABQ decrease gait stability with significant higher acceleration values. In case of patients with high FABQ and low HAD, higher peaks of propulsion are noticed. After FRP, none significant particular motor control strategies are observed. These results need a specific discussion.

Considering difference on preferred walking speed as a reasonable indicator of walking capacity benefit, the results support the hypothesis that the FRP tend to reveal higher capacity benefit only in the case of patients with high FABQ score and low HAD score. This result is associated to exaggerate peaks of propulsion for this subgroup. This result tends to support our hypothesis in relation to a particular motor control associated to high fear-avoidance but low anxiety and depression symptoms. Lee et al. indicated no significant correlations between FABQ scale and self-selected speed.¹⁵ So, linear correlation does not seem adapted in the case of patients with cLBP. This result could be in relation with Demirel et al. results.¹⁴ Our results show that capacity benefit presents higher clinical interest in the case of studies based on fear-avoidance. Moreover, in line with Demirel et al. who studied disability, we depict that patients with cLBP and severe or moderate biopsychosocial consequences (evaluated with fear-avoidance, anxiety and depression) had different ways to normalize their gait. Whereas FRP can reduce significantly fear-avoidance, anxiety, depression and pain, no significant differences appears on these ways after FRP.

Regarding GRF, only peaks extracted on vertical axis in the second part of the stance phase revealed a significant difference between our three groups. In the case of patients with cLBP and asymptomatic subjects, Yazdani et al. studied foot pressure during gait.²² In this study, patients with cLBP demonstrated greater ground reaction force and impulse in the 3–5th metatarsals mask than the control group.²² This specific impulse characterized by pressure system could be compared to our results from GRF. However, Yazdani et al. did not evaluate the relation between this protective mechanism in function of Fear-Avoidance Model (with fear-avoidance beliefs, anxiety or depression). Thus, our results suppose a specific protective mechanism independently of walking speed, as developed by Yazdani et al.²² This mechanism corresponds to a motor control adaptation applied in function of Fear-Avoidance and anxiety/depression. Exaggerate peak of propulsion could be considered as a simple biomechanics indicator of high fear-avoidance belief associated to low level of anxiety and depression.

Assessment of gait patterns during normal overground walking is typically used for a better understanding of the postural control system and motor control unlike other studies.⁶ Our results reveal lower gait stability of the Center of Mass on vertical axis in case of low fear-avoidance. Gait stability computed on CoM on vertical axis could be considered as a simple biomechanics indicator of low fear-avoidance beliefs.

The difference between the LHAD and HHAD groups would support and clarify the model of Vlayen et al.¹⁰: fears of movement would evolve over time with the appearance of an anxiety or depression state after the “fear of movement” stage. Even if patients describe fear-avoidance beliefs (evaluated with a higher FABQ score), no specific beliefs were adopted. In case of HHAD group, the movement dynamics would be identical to those observed in patients without a high fear of movement. This group could be related to resilience with three points: a positive way on performance (none high peak of propulsion as developed by the LHAD group), a positive way on motor control (low acceleration values on CoM associated to gait stability as developed by the LFABQ group) but a negative impact with high risk of anxiety/depression. This group is sensitive to FRP with significant lower FABQ, HAD and VAS scores after this program.

Originally proposed by Lethem et al.,⁸ the Fear-Avoidance Model of Musculoskeletal pain is based on pain catastrophizing, fear of pain or fear of movement, and pain anxiety. The original FABQ includes two subscales: FABQw and FABQpa. Moreover, even if the psychometric properties of the subscales are better established than the total FABQ, it seems difficult to dispense specific subscale according to patients who work or not for instance, while the questionnaires are usually handed out to the patient group. In addition, in the case of patients who do not work, identifying their fears related to their return to work or future work is a follow-up element during a rehabilitation program. For this reason, total FABQ score was considered in this study.

In order to improve the clinical interest of the FAM, identifying subgroups was introduced.² We considered three subgroups of cLBP patients. FABQ score revealed different evolutions after FRP according to these groups. Furthermore, in Pakzad et al.‘s approach, two subgroups based on their scores on the PCS had a lower FABQ difference than in our study.² They revealed strong correlations between PCS and FABQ scores, and between PCS and trunk muscle activity. Moreover, patients with high FABQ score showed a significant higher EMG activity on specific muscles.² This muscular activity could be related to behaviors associated to hypervigilance and avoiding or escaping activities. This higher muscular activity, associated to high PCS and higher FABQ score, could be linked with higher peaks of propulsion quantified in our study concerning the LHAD group.

We note four major limitations. Firstly, this pilot study is based on a limited size with a limited number of strides explored. Fifteen patients per subgroup and five strides per patients were analyzed. Moreover, this pilot study introduces a new approach with three subgroups contrary to other studies.^2,6 Secondly, we considered Vlaeyen and Linton's model according anxiety and depression could appear in high fear-avoidance beliefs. We did not specifically study the HAD of the LFABQ group. Thirdly, quantified parameters were obtained from gait analysis. The conditions under which gait data were collected could influence the outcome measures.²⁴ Patients with cLBP showed fear-avoidance behaviors, which minimized excessive lumbosacral movements resulting in a more uniform walk despite changes in external stimuli.²³ Nevertheless, we adopted a similar protocol before and after FRP. So, this effect seems to be minimized. Finally, the LHAD group was mainly composed of women. In relation to upper body accelerations, attenuation strategies are different according to gender and age, particularly on pelvic and head.^16,25 These attenuation strategies could influence GRF and gait stability.

5. Conclusion

The aim of this study was to demonstrate the influence of Fear-Avoidance Beliefs, anxiety and depression on dynamic parameters and gait stability for patients with cLBP before and after FRP. To our knowledge, this is the first study which considers three groups of patients with cLBP according to FABQ and HAD scores and details the relation between FAM and biomechanics parameters for patients with cLBP treated with McKenzie procedure. With a limit related to the sample size per group, the results of this study suggest to add new exercises (based on motor control) or treatments according to HAD and FABQ scores, to analyze the links between the biopsychosocial scores and the dynamic or motor control parameters.

Funding

None.

Ethical approval

Ethics Committee of Angers (n° 2021–064).

Declaration of competing interest

None.