CLBP subjects were recruited in a local community through posted advertisements. Inclusion criteria were pain in the lumbar and sacroiliac joint regions that had lasted more than 12 weeks before the test and an average pain intensity score of ≥2/10 on the numerical rating scale (NRS) in the past week. Alshehre et al. (12) found a moderate effect for the Y balance test in young adults with CLBP of NRS ≥ 2/10 (12). Meanwhile, exclusion criteria were patients who reported current pregnancy, signs of nerve root compression, serious spinal conditions (e.g., tumor, infection, or fracture), specific balance issues (e.g., vestibular apparatus problem or uncorrected visual disturbances), history of surgery to the lower extremity or spine, or pain in the joints of the lower extremities. In addition, none of the participants had been treated with CLBP in the past 12 weeks. This study was approved by the Ethics Committee of Zhengzhou Central Hospital Affiliated to Zhengzhou University (No.: 201966).

Once participants were determined to be eligible for the study, their demographic and clinical characteristics were collected, including age, sex, height, weight, and body mass index (BMI), as well as leg length (LL), duration, and location of LBP. LL is the distance between the anterior superior iliac spine and the ipsilateral medial malleolus at supine position (12, 21). Furthermore, participants rated their current pain and average pain intensity in the past week using the NRS. Participants also completed the Roland Morris Disability Questionnaire (RMDQ) to evaluate LBP-related disability, the Self-Rating Anxiety Scale (SAS) to evaluate anxiety in the last week, the Self-Rating Depression Scale (SDS) to evaluate depression in the last week, and the Fear-Avoidance Beliefs Questionnaire for physical activities (FABQp) and work activities (FABQw) to evaluate the fear-avoidance level.

As a widely employed scale with good validity and reliability, RMDQ can allow quantification and further classification of LBP by the extent of its impact on physical function (19). Although cutoff points of RMDQ have not been established clearly, Kimachi et al. (19) found that the level of dysfunction based on RMDQ was associated with an increased risk of falls in the elderly, and the incidence of any fall over the past year was 27.8% of those with low LBP-related disability (RMDQ = 1–5) and 37.6% of those with medium to high level of LBP-related disability (RMDQ = 6–24) (19). Thus, in the present study, participants were divided into two subgroups based on the degree of LBP-related disability measured by RMDQ to compare the difference in postural balance: low (RMDQ = 1–5) and medium to high LBP-related disability (RMDQ = 6–24).

Herein, subjects were asked to perform the OLS to assess their static postural balance (22). During the test, subjects were asked to stand on one leg on a force platform with bare feet and eyes closed. The test ended when the non-standing foot touched the floor or the standing leg, and a timer was used to record the duration. The order of one-leg stance on the left (OLS-L) and right (OLS-R) legs was randomized with random numbers table to minimize the order effect. Subjects practiced thrice on each leg for familiarization. Then, they completed three trials for each leg with the mean value for analysis. There was a 30 s break between trials to avoid fatigue.

Afterward, subjects were instructed to complete YBT correctly to evaluate their dynamic balance (12). The test involves maximal reaching distances in anterior (AT), posterolateral (PL), and posteromedial (PM) directions from the central stance platform at an angle of 120° (Figure 1). During the test, participants were instructed to keep the standing leg on the central stance platform barefooted, and use another bare foot to slide and reach the indicators in three directions as far as possible, and then return to their initial standing position while maintaining their balance. Moreover, the testing orders of the reaching directions and legs were randomized with random numbers table to minimize the order effect. Subjects practiced thrice in each direction for familiarization. Then, they completed three trials in each direction with the mean value for analysis. There were 10 s breaks between trials and 30 s breaks between reach directions and between legs to minimize fatigue. The outcome measures of interest were the distances to be reached in the three directions, the composite score (CS), and the normalized reach. The composite score was calculated as [(AT ± PM ± PL)/(3 × LL)]. Normalized reach was calculated as [(reach distance/LL) × 100%] (12).

According to Alshehre et al. (12), a calculation by G*Power software (v3.1) showed that at least 62 participants (31 per group) were required to detect a moderate effect of Y balance test (power = 0.80, α = 0.05) (12). All data were analyzed using the SPSS software (v23.0). The statistical significance level was set at p < 0.05. The Chi-square test was utilized to explore the significant differences in categorical variables between the groups. The Kolmogorov–Smirnov test was used to analyze whether or not the continuous variables were normally distributed. To test the first hypothesis, independent T-tests (for normally distributed variables) and Mann-Whiney U-tests (for non-normally distributed variables) were utilized to explore the balance performance between low and medium to high LBP-related disability groups. To address the second hypothesis, Spearman correlations were applied to describe the relationships between postural balance and its influencing factors.

A total of 49 participants with low LBP-related disabilities and 33 participants with medium to high LBP-related disabilities participated in this study. Participants’ characteristics of both groups are summarized in Table 1. The low LBP-related disability group consisted of 19 men and 30 women with a mean age of 26.16 ± 5.47 years, whereas the medium to high LBP-related disability group consisted of 11 men and 22 women with a mean age of 28.42 ± 6.50 years. However, demographic characteristics of age, gender, height, weight, BMI, and LL were not significantly different between groups (p > 0.05). Moreover, no significant clinical differences were observed in LBP duration, pain location, pain intensity, and anxiety between groups (p > 0.05). However, significant differences were found in depression and fear-avoidance levels (including total score, FABQp, and FABQw), which were higher in the medium to high LBP-related disability group (p < 0.05).

Table 2 displays the OLS scores for both groups. The OLS-L score was significantly higher in the low LBP-related disability group (z = −2.081, p = 0.037), whereas the OLS-R had no significant difference between groups (t = 1.429, p = 0.157).

Table 3 presents the YBT scores for both groups. No significant differences were found in the absolute reaching distances in all directions and composite scores between groups. However, concerning the normalized reach, the low LBP-related disability group had significant scores in PM (p = 0.038) and CS (p = 0.026) of the left leg reach, PM (p = 0.032), PL (p = 0.036), and CS (p = 0.027) of the right leg reach as compared to those of the medium to high LBP-related disability group.

Table 4 shows the relationships between balance and demographic and clinical characteristics. The normalized values of left and right leg reach were selected for analysis. Age was negatively correlated with both legs’ reach in multiple directions and OLS-R score. Duration of LBP was positively correlated with left leg reach in the AT direction (r = 0.287, p = 0.005). As for the pain intensity, current pain is positively related to the left leg reach in the AT direction (r = 0.228, p = 0.020). SAS showed an inverse correlation with left leg reach in the PM direction (r = −0.186, p = 0.047) and CS (r = −0.223, p = 0.022), and SDS was negatively related to left leg reach in the PL direction (r = −0.193, p = 0.041) and CS (r = −0.208, p = 0.030), and right leg reach in the AT direction (r = −0.205, p = 0.032), and CS (r = −0.210, p = 0.029). Moreover, fear-avoidance levels (including total score, FABQp, and FABQw) were negatively correlated with both legs reaching in multiple directions, and FABQp presented an inverse correlation with the OLS-R score (r = −0.215, p = 0.026).

In our study, CLBP patients with medium to high dysfunctions were less likely to maintain their balance in OLS-L as compared to those with low dysfunctions, thereby demonstrating a different outcome to the findings of Tsigkanos et al. (23) which indicates that CLBP participants had worse performance in the OLS-R (23). It could be because of differences in the included subjects. Tsigkanos et al. (23) included patients with a higher mean age and no assessment of dysfunction. Thus, the finding of OLS partially supports the original hypothesis that the more severe the LBP-related disability, the worse the static balance.

Although no significant differences were observed in the absolute values of the YBT reaching distance between groups, the normalized reach distances in PM, PL, and CS were significantly higher in the low LBP-related disability group as compared to the medium to high LBP-related disability group. This may be because highly dysfunctional subjects may be accompanied by functional impairments in other body parts, such as lower limbs. For example, poor balance in the posterior direction predicted ankle sprains (24, 25). The YBT composite score showed that the CLBP participants’ ability to perform multiplanar motion decreased with the increase of the dysfunction (26). However, no differences were observed in the AT direction, which was similar to the findings of Hooper et al. (2016) (21). It may be due to the vision that compensates for any somatosensory and vestibular issues during the reaching in AT direction (21, 27). In PM and PL directions, the lower limbs are placed out of sight behind the body, thereby relying more on somatosensory and vestibular sense to complete the balance tasks. That is, the increase in difficulty may further expose deficiencies in dynamic balance in CLBP patients with higher dysfunction, again supporting our original hypothesis.

Perfect posture balance requires stable nervous system function. Normal nervous functions base on normal anatomical structures such as cerebellum, spinal cord, and peripheral nerves, which are essential for the processing of balance information and the timely and effective balance response. Although this paper cannot provide evidence of neural mechanisms, CLBP has been shown to cause neurological changes, such as cerebellar atrophy, white matter damage, and decreased sensory perception (9). And postural balance is related to functional parameters such as muscle strength, flexibility, muscle control, and proprioception (23). Thus, postural balance impairments in CLBP patients implicate not only abnormal lumbar muscle conditions but also underlying neurological changes leading to neural inefficiency, which can be further investigated by combining neuroimaging techniques such as functional near-infrared spectroscopy.

Our findings of a significant negative correlation between age and performance of YBT and OLS are consistent with previous research (22, 23). Postural balance deteriorates with age possibly caused by biological changes (e.g., neuromusculoskeletal alterations, mobility issues, and sensory-motor deficits) (22, 28). Particularly, impaired postural balance increases the risk of falls in older CLBP patients (29). In addition, BMI is unlikely to affect postural balance, which is consistent with da Silva et al. (2019) (22). Interestingly, LBP duration and current pain intensity were found to be significantly positively correlated with the left leg reaching distance (normalized value) in AT direction. However, Brech et al. (4) found that the lumbar pain intensity and frequency were not correlated with postural balance in women with CLBP (4), which may be because the included participants were younger and their compensatory capacity was high (30). Moreover, we found that the scores of depression and fear avoidance belief (except for anxiety) were higher in the medium to high LBP-related disability group than in the low LBP-related disability group. Moreover, anxiety, depression, and fear avoidance belief were negatively related to YBT performance in multiple directions and only the FABQp score presented a significant inverse correlation with the OLS-R score. Recently, researchers have proposed that negative emotions could affect motor control. For example, anxiety and depression disrupt sensory integration and cognitive interaction (31, 32), requiring greater attention demands to maintain balance. However, although there is a lack of evidence on the effect of pain-related anxiety and depression on postural balance, it is theoretically possible that anxiety and depression could affect postural balance (16). Furthermore, CLBP patients with high fear-avoidance beliefs were not only vigilant but also significantly restricted their range of motion during lumbar flexion and extension (18). Fear may also disrupt the automaticity of neural control pathways, leading to deficits in trunk motor control and increased trunk variability during maintaining balance, which may also increase the risk of LBP (33). Dynamic balance is a more difficult condition than static balance, which makes it easier to reveal the relationship between negative emotions and dynamic balance ability in YBT (34).

However, there are still some limitations despite the findings in this paper. First, we included only young subjects and did not include subjects from other age groups, such as elderly patients with CLBP. And we lack evidence from healthy controls, i.e., healthy young adults with no disabilities. Thus, future studies should study a wider age range of subjects and include healthy controls to enhance the applicability of the findings in this paper. Second, this is a cross-sectional study. Thus, longitudinal research can be further carried out to investigate the effects of LBP-related disability and other factors on postural balance. Furthermore, this paper is limited to behavioral results and lacks neuroimaging evidence, so neuroimaging techniques should be further used to explore the neural mechanism of the effect of LBP-related disability on balance impairments.