Background

Low back pain (LBP) is a prevalent musculoskeletal disorder, impacting approximately 70% of the adult population in the world1, resulting in significant treatment costs and absence from work time2. 90% of LBP cases cannot be ascertained through objective examination, as they lack an identifiable underlying pathology and are defined as nonspecific low back pain (NSLBP)3,4. Multimodal treatment has been widely utilized in clinical practice, attributable to the increasing occurrence of NSLBP in recent years5. Among the various treatments, exercise therapy is the cornerstone of treatment for musculoskeletal conditions and is advised for the rehabilitation of NSLBP patients6.

A systematic review indicates that chronic LBP is associated with reduced postural control7. In contrast to healthy individuals, those with LBP tend to restrict trunk movement to alleviate discomfort; however, this restriction adversely affects lumbar proprioception8,9and raises the perceptual threshold10. Moreover, people with NSLBP generally demonstrate a delayed activation of trunk muscles, undermining lumbar spine stability and diminishing postural control and balance11.

Hamstring tightness is a risk factor for LBP12. Certain patients with LBP are accompanied by shortened hamstrings12, and those experiencing recurrent LBP showed considerably impaired hamstring flexibility13. Shortened hamstrings can lead to a flat back posture, elevating lumbar disc pressure and requiring compensation from the lumbar muscles14. Ultimately, this compensatory process may lead to symptoms of LBP.

Exercise therapy is a significant treatment for NSLBP, mitigating the adverse effects of pharmacological interventions, and is extensively utilized in rehabilitation science fields15. Balance training can enhance trunk muscle recruitment and neuromuscular control by addressing delayed muscle activation and inadequate proprioception16. Current research indicates that trunk balance combined with flexibility training is more effective than strength combined with flexibility training in the recovery of disability and quality of life16. Furthermore, improving postural compensation caused by hamstring tightness, hamstring training can release the overstrained hamstring muscles, decrease pressure in the lumbar intervertebral discs, and lessen the load on the lumbar spine13, consequently promoting the stability of lumbar vertebrae and pelvis17. In clinical settings, studies showed that increasing hamstring flexibility effectively reduces pain intensity, increases lumbar range of motion, and decreases disability12,13.

The advantages of balance and hamstring training in patients with LBP have been separately established. However, current experimental research encounters obstacles, including intervention heterogeneity and nonspecific methodologies, which limit their practical application. This study is the initial study to integrate balance and hamstring training for the management of NSLBP, employing a more comprehensive training programme. Balance training comprises dynamic and static balance exercises for an overall regimen, while hamstring training incorporates both stretching and strength-building exercises aimed at the hamstring muscles. We implemented a 6-week training programme to examine the efficacy of this combination training in improving individuals with NSLBP.

Materials and methods

Setting of participants, treatment, and data collection

Participant recruitment was carried out at Shenzhen University, Shenzhen City, China. The treatment sessions consisted of a 6-week exercise intervention conducted in the laboratory of the School of Physical Education, Shenzhen University. The sample size was calculated using G*Power software (Version 3.1.9.7, Franz Faul, University of Kiel, Kiel, Germany)18. A sample size of 24 was determined to achieve a significance level of 0.05, a power of 0.80, and an effect size of 0.30. A sports therapist provided personalised guidance for participants. Self-reported scales and physical assessment data from participants were collected by the same assessor at baseline and the week following the completion of the intervention. This prospective intervention study aims to recruit patients with NSLBP. This study adhered to the Consolidated Standards of Reporting Trials (CONSORT) guidelines19. All participants in the study provided informed consent. This study was approved by the Medical Ethics Committee, Department of Medicine, Shenzhen University (Ethical approval number: PN-202200024) (Date of ethical approval: 07/06/2022). This trial was registered in the International Standard Registered Clinical/Social Study Number (ISRCTN) registry (Trial registration number: ISRCTN14488937) (Date of trial registration: 28/05/2024). All methods in this study complied with the relevant guidelines and regulations established by the International Standard Registered Clinical/Social Study and Medical Ethics Committee of Shenzhen University.

Eligibility criteria

Inclusion criteria

Participants were included according to the following criteria: (1) aged 18–65 years old (to avoid confounding factors due to older age); (2) have NSLBP (defined as pain perceived anywhere in the 12th costal to gluteal fold in the past year, with or without numbness and radiating pain in the lower extremities); (3) have a normal cognitive function, with no history of craniocerebral injury, cerebrovascular disease, epilepsy, or other complication;(4) can attend moderate physical activity; (5) have not participated in any LBP-related exercise rehabilitations in the three months prior to the study; (6) have a balance test score of at least 1.5; (7)must give their informed consent to this study.

Exclusion criteria

The exclusion criteria were as follows: participants will be excluded from this study (1) if their LBP is associated with specific pathoanatomical causes (e.g., tumor, tuberculosis, fracture, or definite infection); (2) if they present with sciatica or radicular pain syndrome; (3) if they have a definite history of spinal surgery (4) if they have severe heart disease and visceral disease; (5) if their NSLBP is in the acute phase.

Interventions

The experimental group (EG) underwent 6 weeks of balance combined with hamstring training, while the control group (CG) did not receive any intervention. During this 6-week programme, participants received personalised guidance and supervision from an experienced sports therapist in 45-minute sessions conducted three times weekly. The training session consisted of four components: a warm-up lasting 5 min, balance training for 20 min, hamstring training for 15 min, and relaxation for 5 min.

The objective of the warm-up was to engage the lumbar and abdominal muscles through basic gymnastic exercise. Balance training incorporated both static and dynamic balance exercises utilizing the TOGU Balance training system (Balance Challenge Disc 2.0 (Balance Trainer) (Fig. 1) and Balanza Components, Germany (Fig. 2)). Participants in balance training were instructed to activate their abdominal muscles and engage their glutes and thighs to maintain core stability. Hamstring training encompasses both stretching and strength training, aimed at alleviating tension in overstrained hamstrings and enhancing muscle strength. Relaxation exercises included abdominal stretching, back stretching, and muscle release techniques using a roller to facilitate muscle relaxation and promote physical recovery. Further details of the intervention programme are presented in Supplementary file 1. None of the participants received any additional treatments for LBP, including medication.

Fig. 1
figure 1

Balance Challenge Disc 2.0. (Image source: TOGU Official Website, https://www.togu.de/en/shop/balance-trainer/315/challenge-disc-2.0)

Full size image
Fig. 2
figure 2

Balanza Components. (Image source: TOGU Official Website, https://www.togu.de/shop/balance-board/)

Full size image

Outcome measurements

Participants were evaluated for pain intensity, balance ability, lumbar muscle functions (including abdominal strength, back strength, abdominal endurance, and back endurance), and hamstring functions (specifically hamstring flexibility and hamstring strength) at baseline and following a 6-week intervention by an independent assessor.

Primary outcome

Pain intensity.

In this study, a 10 cm Visual Analogue Scale (VAS) was employed to evaluate pain intensity, where 0 cm indicated “no pain” and 10 cm denoted “the worst pain imaginable”. Participants selected a numerical value on the scale to represent their intensity of LBP20. The assessor in this study is skilled in using the VAS scale correctly and has extensive experience with its application in previous studies. All participants used the VAS scale under the guidance of the assessor.

The secondary outcomes

  1. (1)

    Balance ability was measured using the TOGU Balance Test system (Balance Challenge Disc 2.0, Togu, Germany)21.

  2. (2)

    Lumbar muscle functions:

Lumbar strength and endurance were evaluated as they are critical indicators of lumbar function and effectively reflect the impact of the intervention.

Abdominal and back strength were measured using an isometric force testing device to determine maximal isometric strength for trunk extension and flexion (Back Check, Dr. Wolff, Germany)22.

Abdominal and back endurance were measured by the maximum sustain time23,24.

  1. (3)

    Hamstring functions:

Hamstring flexibility was assessed using the sit-and-reach test field25.

Hamstring strength was measured using the isometric force testing device (Back Check, Dr. Wolff, Germany).

Further details regarding the outcomes and measurements are presented in Supplementary file 2.

Statistical analysis

The pre-test data were analyzed through an independent sample t-test following an assessment of data normality. A two-way repeated measures analysis of variance (ANOVA) was conducted on the post-test data to examine the main effects of group and time, along with the interaction effect of time × group. The interaction effect of time and group was subsequently analyzed through simple effects analysis. Prior to conducting any statistical analyses, the normality of the data was evaluated.

In cases where the data failed to satisfy the assumption of normality, suitable non-parametric tests, such as the Mann-Whitney U test for independent samples, the Wilcoxon signed-rank test for paired samples, or the Friedman test for repeated measures, were employed as alternatives. All analyses were conducted using SPSS 26.0, with data presented as mean (standard deviation) (MSD).

Participants and procedure

Thirty-seven participants were recruited via the Internet. After excluding ineligible participants, 26 participants were considered eligible for the study (18 males and 6 females; mean age: 25.73 ± 7.74 years). Participants were randomly allocated to either the EG or the CG using computer-generated random numbers, with 13 participants per group. At baseline, all participants were evaluated for pain intensity (VAS), balance ability (TOGU balance test), lumbar muscle functions (strength and endurance), and hamstring functions (extensibility and strength). Baseline differences between the groups were analyzed using an independent samples t-test.

Results

Demographic characteristics and baseline assessment

The results indicated no significant differences between the EG and the CG in the VAS score, balance ability, lumbar muscle functions, or hamstring functions at the baseline (p > 0.05). Baseline characteristics and statistical outcomes are summarized in Table 1.

Table 1 Data of baseline assessment (n = 26).
Full size table

After six weeks of training, 24 participants completed all outcome measures. Two participants provided pain intensity (VAS score) but did not attend physical assessments.

Changes in primary outcome (pain intensity) before and after training in both groups

A combination of the Wilcoxon Signed-Ranks Test and Mann-Whitney U Test was employed to assess changes in pain intensity (pre- and post-training) between the EG and the CG. The analysis revealed no significant difference in baseline pain intensity between EG and CG prior to the intervention (U = 81.500, Z = −0.172, p = 0.863) (Table 2). However, a significant reduction in pain intensity was observed within EG following the intervention (Z = −3.246, p = 0.001), while no significant change was observed in the CG (Z = −1.342, p = 0.180). Furthermore, the reduction in pain intensity in EG was more significant than in CG (U = 0.000, Z = −4.457, p < 0.001). These findings indicate a marked decrease in VAS scores following the balance and hamstring training for people with NSLBP (Table 3).

Table 2 Main and interaction effects of group and time on outcomes.
Full size table
Table 3 Intra- and inter-group differences in outcomes before and after intervention.
Full size table

Changes in the balance ability before and after training in both groups

The analysis revealed that the main effect of group was not significant (F(1, 22) = 0.088, p = 0.770, η²p= 0.004). However, the main effect of time was significant (F(1, 22) = 8.707, p = 0.007, η²p= 0.284), and the interaction between group and time was also significant (F(1, 22) = 8.833, p = 0.007, η²p = 0.284) (Table 2). Further simple effects analysis revealed the following: (1) There was no significant difference in balance ability between the EG and the CG either pre- or post-training (p> 0.05): (2) Compared with the baseline, the balance ability of the EG was improved significantly following the intervention (p < 0.001), whereas no significant changes were observed in the CG (p > 0.05) (Table 3).

Changes in lumbar muscle function before and after training in both groups

Abdominal strength

The results showed that the main effect of group was significant (F(1, 22) = 4.788, p = 0.040, η²p= 0.179). Similarly, the main effect of time was also significant (F(1, 22) = 14.034, p = 0.001, η²p= 0.389). However, the interaction effect of group × time was not significant (F(1, 22) = 0.541, p = 0.470, η²p = 0.024) (Table 2). Further simple effects analysis revealed the following: (1) There was no significant difference in abdominal strength between the EG and the CG, either before or after the training (p≥ 0.05); and(2) Compared with the baseline, abdominal strength in the EG showed a significant improvement following the training (p = 0.003), while no significant change was observed in the CG (p > 0.05) (Table 3).

Back strength

The analysis revealed that the main effect of group was significant (F(1, 22) = 3.534, p = 0.0073, η²p= 0.138). Similarly, the main effect of time was also significant (F(1, 22) = 18.277, p < 0.001, η²p= 0.454). However, the interaction effect of group × time was not significant (F(1, 22) = 3.995, p = 0.058, η²p = 0.1544) (Table 2). Further simple effects analysis showed the following: (1) There was no significant difference in back strength between the EG and the CG before training (p > 0.05), but a significant difference was observed between the two groups after training (p = 0.015). (2) Compared to baseline, back strength in the EG improved significantly after the intervention (p < 0.001), whereas no significant changes were found in the CG (p > 0.05) (Table 3).

Abdominal endurance

The results showed that the main effect of group was not significant (F(1, 22) = 3.036, p = 0.095, η²p= 0.121), while the main effect of time was significant (F(1, 22) = 33.002, p < 0.001, η²p = 0.600). Additionally, the interaction effect of group × time was significant (F(1, 22) = 7.384, p = 0.013, η²p = 0.251) (Table 2). Further simple effects analysis revealed the following: (1) There was no significant difference in abdominal endurance between the EG and the CG before training (p > 0.05), but a significant difference was observed between the two groups after training (p = 0.032). (2) Compared to baseline, abdominal endurance in the EG improved significantly after training (p < 0.001), whereas no significant changes were observed in the CG (p > 0.05) (Table 3).

Back endurance

The results indicated that the main effect of group was not significant (F(1, 22) = 0.543, p = 0.469, η²p = 0.024), while the main effect of time was significant (F(1, 22) = 20.847, p < 0.001, η²p = 0.487). Additionally, the interaction effect of group × time was significant (F(1, 22) = 18.394, p < 0.001, η²p = 0.455) (Table 2). Further simple effect analysis revealed the following: (1) There was no significant difference between the two groups before training (p > 0.05), but a significant difference was observed between the two groups after training (p = 0.027). (2) Compared with the baseline, back endurance in the EG improved significantly following the intervention (p < 0.001), while no significant changes were observed in the CG (p > 0.05) (Table 3).

Changes in hamstring functions before and after treatment in both groups

Hamstring flexibility

The results indicated that the main effect of the group was not significant (F(1, 22) = 0.009, p = 0.926, η²p< 0.001), while the main effect of time was significant (F(1, 22) = 11.227, p = 0.003, η²p= 0.338). However, the interaction effect of group × time was not significant (F(1, 22) = 1.825, p = 0.190, η²p = 0.077) (Table 2). Further simple effects analysis revealed the following: (1) There was no significant difference in hamstring flexibility between the EG and the CG either before or after the training (p > 0.05);  (2) Compared to baseline, hamstring flexibility in the EG showed a significant improvement following the intervention (p = 0.002), whereas no significant changes were observed in the CG (p > 0.05) (Table 3).

Hamstring strength

The results indicated that the main effect of the group was significant for both the left side (F(1, 22) = 6.437, p = 0.019, η²p= 0.226) and the right side (F(1, 22) = 6.080, p = 0.022, η²p = 0.217). The main effect of time was also significant for both the left side (F(1, 22) = 13.679, p = 0.001, η²p= 0.383) and the right side (F(1, 22) = 17.669, p < 0.001, η²p = 0.445). Additionally, the interaction effect of group × time was significant for both the left side (F(1, 22) = 4.359, p = 0.049, η²p = 0.165) and the right side (F(1, 22) = 7.953, p = 0.010, η²p = 0.266) (Table 2). Further simple effects analysis revealed the following: (1) There were no significant differences in hamstring strength between the EG and CG before training (p > 0.05), but significant differences were observed between the two groups after training for both the left side (p = 0.007) and the right side (p = 0.005). (2) Compared to the pre-test, hamstring strength in the EG improved significantly after training for both the left and right sides (p < 0.001), whereas no significant changes were observed in the CG for either side (p > 0.05) (Table 3).

Discussion

The findings demonstrated that for the primary outcome, six weeks of balance and hamstring training significantly reduced pain intensity in participants with NSLBP compared to those who did not receive any intervention. For the secondary outcomes, the training significantly improved back strength, abdominal endurance, back endurance, and hamstring strength relative to no intervention. However, no significant improvements were observed in balance ability, abdominal strength, or hamstring flexibility in participants with NSLBP compared to those who did not receive any intervention.

How balance and hamstring training reduce pain intensity

This study indicated that balance and hamstring training significantly reduced pain intensity in individuals with NSLBP. Participants who underwent this training showed a significant enhancement relative to those who did not receive any intervention.

Exercise is a crucial element in rehabilitating chronic musculoskeletal conditions26. Initially, exercise has been demonstrated to correspond with alterations in pain perception (hypoalgesia)27. During physical activity, increased blood pressure is inversely correlated with pain perception27, suggesting that individuals with back pain may experience reduced pain sensitivity as their blood pressure rises during exercise27. Secondly, exercise enhances pain thresholds by stimulating the production of endogenous opioids, which naturally act to reduce pain perception, therefore relieving LBP28. In addition, exercise also plays a role in regulating inflammatory factors. Research has indicated that exercise contributes to pain relief by reducing muscle inflammatory response and enhancing descending pain inhibition29,30. Pain relief is also a cumulative result of improved muscle strength and endurance. By engaging in 6-week training programmes, individuals with NSLBP can experience a significant recovery in back strength, abdominal endurance, and back endurance, which positively protect the lumbar joints and contribute to better posture control. These improvements in lumbar health, in turn, reduce the intensity and recurrence of LBP.

Studies support the association between balance or hamstring training and pain reduction. Dae-Hyun et al. reported that eight weeks of balance training significantly reduced pain intensity in women with chronic LBP31. Similarly, Ju-hyun et al. indicated that hamstring stretch therapy effectively decreased pain levels in individuals with radicular LBP32. These findings respectively highlight the significance of balance and hamstring training in controlling pain intensity for LBP.

How balance and hamstring training improve balance ability

Although our results indicate no significant differences in the post-test between the EG and CG groups, our findings still support the potential advantages of improving balance ability for people with NSLBP, as evidenced by the significant increase in balance scores of the EG after training.

The TOGU balance training system, employed in this study, incorporates a balance challenge disc to assess participants’ balance ability and provides training at various difficulty levels. Furthermore, this system enables participants to perform various balance training exercises using the Balanza components, which provide an unstable surface. Training on such unstable surfaces requires participants to continuously transition between static and dynamic balance while maintaining body posture control. This process improves proprioceptive sensitivity, neuromuscular control, and motor coordination in the lumbar region, which collectively contribute to postural balance improvements in patients with NSLBP33.

David et al. supported the use of unstable devices in training patients with LBP34. This study indicated that, for the rehabilitation of individuals with LBP, in addition to strengthening the lumbar core muscles, it is also necessary to improve the stability and endurance of these muscles. When resistance training is applied on unstable surfaces, core muscle activation and neuromuscular coordination can be effectively enhanced, which directly contributes to the stability of joints in the spinal region34.

The insignificant differences between groups may be attributed to a short training period and/or not individualizing the training. However, the positive impacts of balance and hamstring training should not be overlooked, given the findings from previous studies and the observed improvements in the training group.

How do balance and hamstring training improve lumbar functions

A study showed that NSLBP patients tended to have delayed activation of the low back muscles compared with healthy people16. Delayed multifidus activation can reduce lumbar stability during changes in body posture, and lumbar instability, in turn, aggravates LBP symptoms35. During balance training, the body is in an unstable support state, requiring the lumbar muscles to be mobilized to maintain the balance36. By improving the coordination and recruitment ability of the neuromuscular system, balance training strengthens the contraction of the lumbar muscles, particularly the deep muscles (e.g., multifidus) and supporting muscles (e.g., erector spinae), thereby improving lumbar muscle strength and endurance16.

Our results demonstrated that balance and hamstring training significantly improved back strength, and abdominal and back endurance in NSLBP participants compared to those who did not receive any intervention. The balance intervention programmes in this study extensively engaged lumber muscles. Participants performed a series of exercises on unstable TOGU surfaces, requiring dynamic and isometric contractions of the abdominal and back muscles to maintain postural stability. Additionally, performing multiple repetitions or maintaining a load for 30 s per single motion likely contributed to the observed improvements in strength and endurance.

How balance and hamstring training improve hamstring functions

In this study, participants with NSLBP were subjected to hamstring stretching in a sitting position, straight leg lifting with a resistance band, and myofascial release with a roller. However, the results indicated that no statistical difference in hamstring flexibility between the EG and CG groups after a training programme. This may be because the training intensity of this programme (three sessions a week, three sets on each side, each set of 30–60 s) needs to be improved or the training period needs to be prolonged.

We suggest that hamstring flexibility should be considered an integral part of treatment for individuals with LBP, as previous studies have identified hamstring tightness as a significant risk factor for LBP12,13. The hamstrings and erector spinae are part of the superficial dorsal line37, and overstressed hamstrings can lead to compensatory adjustments in the function of back muscles within this line37. Shortened hamstrings cause the posterior pelvic tilt, which leads to lumbar instability and muscle imbalance38. Hamstring stretching has been shown to improve poor posture in NSLBP patients, improve lumbar function, and reduce pain.

Our study also showed that the hamstring strength of the NSLBP participants was significantly increased compared to those who did not receive any intervention, which proves that this training is an effective way to strengthen the hamstring muscle. Kasmi et al. conducted 6 weeks of hamstring training on 10 elite rowers and found that a period of training could significantly increase the pain threshold of rowers and moderately improve their sports performance17. This study combined eccentric self-weight with concentric weight-bearing training, strengthened the hamstring contraction, and improved the hamstring weakness that may exist in NSLBP people.

Study innovations and limitations

Previous studies have focused on the separate effects of balance training or hamstring training. As mentioned earlier, Dae-Hyun et al. and Ju-hyun et al. provide supporting evidence for balance or hamstring training at pain relief levels31,32. Targeting the hamstring shortness and postural control weakness of NSLBP people, this study is the first to combine balance and hamstring training for the treatment of NSLBP.

Comprehensive training programmes were employed in this study. For balance training, a combination of static and dynamic balance exercises was utilized. The dynamic balance training included movements performed on dynamic planes, such as deadlifts, skateboard spin squats, and swallow balance exercise, requiring participants to improve their controlled coordination and stability. This innovative balance training represents a key advancement in our intervention strategy, differentiating it from previous studies.

For hamstring training, hamstring stretching is a more common intervention in current studies32,39. In our study, both hamstring stretching and strengthening exercises were included in the training programmes, which comprehensively explored the role of hamstrings in NSLBP rehabilitation. However, there are limitations to the study. Firstly, the relatively small sample size may limit the applicability of our findings. Secondly, the intervention duration of 6 weeks may not have been long enough to observe significant differences between the two groups for balance ability, hamstring flexibility, and abdominal strength. Finally, we only set up a control group due to the resource limitation. In the future, separate training groups should be considered to investigate the exclusive effects of balance or hamstring training for the NSLBP individuals and compare them with the combined approach.

For NSLBP people, the options for exercise intervention are varied, such as whole-body vibration exercise and spinal stabilization exercise. It is worth noting that whole-body vibration exercise proved to be an effective treatment for NSLBP in clinical studies. Systematic reviews show that this treatment benefits the recovery of pain, disability, proprioception, and quality of life relative to functional performance. Whole-body vibration exercise can be a complementary treatment to balance and hamstring training40. Another systematic review illustrates that spinal stabilization exercise has a positive effect on balance improvement for the LBP population; however, low certainty of evidence adds an element of uncertainty to the application of this treatment41.

Conclusion

Six weeks of balance and hamstring training can effectively reduce pain intensity and improve back strength, abdominal endurance, back endurance, and hamstring strength. This treatment could be considered part of NSLBP treatment in future clinical practice.