Responsiveness and minimal clinically important difference of the Brief BESTest 12 months after surgery for lumbar spinal stenosis: a prospective cohort study in Japan
Article information
Abstract
Study Design
Prospective cohort study.
Purpose
To assess the responsiveness of the Brief Balance Evaluation Systems Test (Brief BESTest) after lumbar spinal stenosis (LSS) surgery and to determine its minimal clinically important difference (MCID).
Overview of Literature
LSS leads to balance impairment and increased fall risk, even after surgical decompression. Although postoperative balance improves, the specific subsystems that recover and the clinically meaningful magnitude of the change remain unclear.
Methods
This prospective study included 197 patients aged ≥60 years who underwent decompression or fusion surgery for LSS. Balance was assessed preoperatively and 12 months postoperatively using the Brief BESTest. Responsiveness was quantified using the standardized response mean (SRM). The MCID was determined using receiver operating characteristic (ROC) analysis with the Global Rating of Change as an anchor. Bootstrap resampling was used to evaluate MCID stability. Patients were classified into moderate or severe groups based on preoperative Brief BESTest scores, using 13 points as the previously identified cutoff for increased fall risk in older adults.
Results
The Brief BESTest scores improved significantly in all patients (11.6±5.5 to 14.1±6.2; SRM=0.51). The severe group showed greater responsiveness (SRM=0.64) than the moderate group (SRM=0.38). In the subsystem, the largest improvements were observed in anticipatory postural adjustments and gait stability, whereas stability limits showed little change. ROC analysis identified an MCID of 2 points (area under the curve=0.79), which was consistent across the severity subgroups. Bootstrap validation confirmed the stability of this 2-point MCID.
Conclusions
The Brief BESTest showed moderate responsiveness to balance improvements after LSS surgery, with marked gains in anticipatory adjustments and gait stability. The MCID of the Brief BESTest after LSS surgery is 2 points, providing a practical benchmark for postoperative evaluation regardless of preoperative severity.
Introduction
Lumbar spinal stenosis (LSS) is a common degenerative condition in older adults characterized by intermittent claudication due to lower extremity pain, numbness, and sensory disturbances [1]. Patients with LSS occasionally experience sensory disturbances in the soles of their feet, such as sensations of walking on small pebbles or a thin membrane covering the plantar surface; these symptoms often persist even after surgery [2]. Moreover, persistent plantar sensory disturbances reduce proprioceptive input and are associated with impaired postural balance, decreased gait performance, and an increased risk of falls [3]. A previous study has shown that balance function improves within 6–12 months after lumbar decompression surgery compared to preoperative levels; however, balance impairments persist [4]. Significantly, persistent postoperative balance deficits may increase the risk of falls in daily life, and approximately 30% of patients experience at least one fall within the first year after LSS surgery [5]. These observations highlight the clinical importance of assessing postoperative balance recovery and identifying factors that influence functional improvement in patients with LSS. Reliable and theoretically grounded assessment tools are required to accurately quantify these postoperative balance changes.
The Balance Evaluation Systems Test (BESTest) is commonly used to assess balance function [6]. This comprehensive tool is based on a theoretical motor control model and evaluates six balance subsystems: biomechanical constraints, stability limits/verticality, anticipatory postural adjustments, postural responses, sensory orientation, and gait stability. Impairment of one or more of these systems can result in postural instability and balance deficits. To facilitate clinical use, a shortened version, the Brief BESTest, was developed based on the same theoretical constructs [7]. Each subsystem of the Brief BESTest includes representative tasks, such as the One-Leg Standing Test (OLS), Timed Up and Go Test (TUG), and Functional Reach Test (FRT), which are commonly used to assess balance in patients with LSS [8–10]. A recent study has shown that improvements in Brief BESTest scores are significantly associated with reductions in postoperative pain, disability, and fear of falling in patients with LSS [11].
Although only a few studies have examined postoperative balance in patients with LSS using the Brief BESTest [11], comprehensive evaluations remain scarce. It remains unclear which balance subsystems of the Brief BESTest are most likely to improve after LSS surgery. Further, the magnitude of improvement in the total score that represents the minimal clinically important difference (MCID) has not been established. Therefore, this study primarily aimed to determine the MCID of the Brief BESTest following LSS surgery; the secondary purpose was to examine the responsiveness of the Brief BESTest.
Materials and Methods
Ethics statement
This study prospectively analyzed patients treated for LSS at our tertiary referral center. The study was approved by the hospital’s institutional review board (Ethics Committee of Okayama Rosai Hospital, approval no., 521-3). Written informed consent was obtained from all participants prior to enrollment.
Study population
A total of 241 individuals who consecutively underwent lumbar decompression or fusion surgery for LSS between April 2022 and November 2024 were screened for eligibility.
Patients were eligible for inclusion if they met the following criteria: (1) age ≥60 years; (2) diagnosis of LSS confirmed by both characteristic clinical manifestations and magnetic resonance imaging (MRI) findings; (3) completion of Brief BESTest evaluations preoperatively and at the 12-month postoperative time point; and (4) provision of signed informed consent. The exclusion criteria were: (1) concomitant cervical or thoracic spinal pathology; (2) advanced osteoarthritis of the hip or knee joints; (3) previous or current neurological, pulmonary, cardiac, or cognitive disorders; (4) preoperative Brief BESTest score ≥20; and (5) missing or incomplete dataset. After applying these criteria, 197 patients were enrolled in the final analysis (Fig. 1).
Patient and surgical factors
Demographic variables included patient age, sex, and body mass index. Surgical factors included the procedure type, spinal levels involved (L1/2 to L5/S1), operation time, and estimated intraoperative blood loss. Perioperative complications were prospectively recorded, with a particular focus on major complications, including surgical site infection, hematoma, and cerebrospinal fluid leakage. Neurological symptom patterns were classified as radicular, cauda equina, or mixed, based on preoperative clinical findings.
Schizas classification
LSS severity was determined on axial MRI using the Schizas classification [12], which classifies the degree of dural sac narrowing from type A to D based on the configuration of the nerve roots.
Oswestry Disability Index
The degree of disease-specific impairment in activities of daily living was assessed using the Oswestry Disability Index (ODI) [13], which consists of 10 items, including pain intensity, walking, sitting, standing, sleeping, and social life. Each item is scored on a 6-point scale (0–5), and the total score is converted to a percentage ranging from 0% to 100%, with higher scores indicating greater disability.
Visual Analog Scale
Pain intensity in the lower back and lower limbs was assessed using a 100-mm Visual Analog Scale (VAS), with 0 representing “no pain” and 100 representing the “worst imaginable pain.”
Brief BESTest
Balance ability was evaluated using the Brief BESTest [7], which assesses postural control across six subsystems: (1) biomechanical constraints, (2) stability limits, (3) anticipatory postural adjustments, (4) postural responses, (5) sensory orientation, and (6) gait stability (Fig. 2). Each item is scored from 0 to 3, with higher scores indicating better balance performance. The total score ranges from 0 to 24 points, with lower scores reflecting greater balance impairment. The Brief BESTest was administered by trained physical therapists at the preoperative and 12-month postoperative assessments.
The Brief Balance Evaluation Systems Test consists of six balance subsystems. (A) Biomechanical constraints (hip and trunk lateral strength), (B) Stability limits (functional reach test), (C) Anticipatory postural adjustments (one-leg standing), (D) Postural responses (lateral compensatory stepping correction), (E) Sensory orientation (stance on foam with eyes closed), and (F) Stability in gait (Timed Up and Go test).
Global Rating of Change
The Global Rating of Change (GRC) was assessed using a questionnaire that asked patients, “Compared with before surgery, how has your sense of stability during mobility changed?” Responses were recorded on a 5-point Likert scale (1=much worse, 2=slightly worse, 3=no change, 4=slightly improved, 5=much improved). Patients who selected “slightly improved” or “much improved” were classified into the improvement group for anchor-based analyses.
Statistical analysis
Normality of continuous variables was assessed using the Shapiro-Wilk test, which indicated non-normal distributions for all variables. Patients were classified into moderate or severe balance impairment groups based on the preoperative Brief BESTest cutoff of 13 points [14]. Between-group comparisons were performed using the Mann-Whitney U test for continuous variables and Fisher’s exact test for categorical variables. For all patients and each severity subgroup, the preoperative and 12-month postoperative Brief BESTest total and subsystem scores were compared using the Wilcoxon signed-rank test. Responsiveness at 12 months was evaluated using the standardized response mean (SRM), with values interpreted according to Cohen’s criteria (small: 0.20–0.49, moderate: 0.50–0.79, large: ≥0.80) [15]. The MCID and area under the receiver operating characteristic (ROC) curve (AUC) were determined using ROC analysis, with the GRC as the external anchor. The precision of MCID estimates was examined via 1,000-iteration bootstrap resampling. AUC values of 0.70–0.79, 0.80–0.89, and ≥0.90 were interpreted as acceptable, excellent, and outstanding, respectively [16]. To address the age imbalance between the severity groups, age-adjusted multivariable analysis was performed using the 12-month Brief BESTest score as the dependent variable and age, baseline Brief BESTest score, and preoperative severity group as independent variables. Statistical analyses were conducted using EZR ver. 1.61 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), and two-tailed p-values <0.05 were considered significant [17].
Results
The study included 197 patients, with 97 in the moderate group and 100 in the severe group (Table 1). Patients in the severe group were significantly older than those in the moderate group (76.4±8.0 years vs. 72.4±8.0 years, p<0.001). Preoperative VAS scores for low back pain, leg pain, and leg numbness were comparable; however, preoperative ODI scores were significantly higher in the severe group (50.4%±18.7% vs. 39.1%±17.8%, p<0.001). The distribution of the Schizas classification did not differ significantly between the moderate and severe groups (p=0.256), and no clear association was observed between stenosis morphology and preoperative balance impairment. The distribution of neurological symptom patterns (radicular, cauda equina, and mixed types) did not differ significantly between the moderate and severe groups (p=0.162). Major perioperative complications occurred in 10 patients (5%). These included deep surgical site infection in four patients, hematoma in two patients, and dural injury with cerebrospinal fluid leakage in four patients. The incidence of major perioperative complications was identical in the moderate and severe groups, with five patients (5%) in each group, and no significant between-group difference was observed (p=1.00).
In the age-adjusted multivariable analysis, the baseline Brief BESTest score and age were independently associated with the 12-month Brief BESTest score, whereas the influence of preoperative severity was not significant after adjustment (Table 2).
Changes in the Brief BESTest total and subsystem scores from the preoperative period to 12 months postoperatively are shown in Table 3. For all patients, the total Brief BESTest score improved significantly (11.6±5.5 to 14.1±6.2, p<0.001), with moderate responsiveness (SRM=0.51). All subsystems except stability limits improved significantly, with the greatest responsiveness observed for gait stability (SRM=0.51).
Changes in Brief BESTest subsystem scores from preoperative to 12 months after surgery according to baseline severity
In the moderate group, the total score increased from 16.2±2.2 to 17.7±4.3 (p<0.001, SRM=0.38). Significant improvements were observed in anticipatory postural adjustments, sensory orientation, and gait stability, demonstrating small-to-moderate responsiveness. In contrast, the severe group demonstrated marked improvement in the total score (7.1±3.8 to 10.6±5.6; p<0.001, SRM=0.64). Significant gains were noted in biomechanical constraints, anticipatory postural adjustments, postural responses, sensory orientation, and gait stability. Responsiveness was greatest in gait stability (SRM=0.68) and anticipatory postural adjustments (SRM=0.50), indicating greater postoperative recovery among patients with severe preoperative impairment.
An MCID of 2 points yielded a sensitivity and specificity of 0.76, with an AUC of 0.79 (95% confidence interval [CI], 0.72–0.84) across all patients. In the moderate group, the AUC was 0.78 (95% CI, 0.68–0.89), whereas in the severe group, the same threshold demonstrated excellent discriminative ability (AUC=0.83; 95% CI, 0.74–0.91). Overall, an MCID of 2 points was consistent across the severity categories (Table 4, Fig. 3). Bootstrap validation with 1,000 resamples confirmed the stability of the MCID estimate. An MCID of 2 points remained robust across all patients (AUC=0.79; 95% CI, 2–3), the moderate group (AUC=0.79; 95% CI, 0–3), and the severe group (AUC=0.83; 95% CI, 2–3).
Receiver operating characteristic (ROC) curves for determining the minimal clinically important difference (MCID) of the Brief Balance Evaluation Systems Test (Brief BESTest). (A) ROC analysis for all patients. (B) ROC analysis for the moderate impairment group. (C) ROC analysis for the severe impairment group.
Discussion
This study demonstrated that balance function improved 12 months after LSS surgery, even among patients with severe preoperative impairment. Although previous research has reported postoperative improvements in overall balance performance and gait [11], no study has examined balance subsystems separately, quantified their responsiveness using the SRM, or established the MCID of the Brief BESTest. In this study, anticipatory postural adjustments and gait stability were more responsive than the other Brief BESTest subsystems, and the MCID was consistently identified across preoperative severity levels. These findings highlight the utility of the Brief BESTest for interpreting postoperative balance recovery in patients with LSS. Furthermore, the age-adjusted multivariable analysis showed that the postoperative balance level was primarily explained by age and baseline balance ability. In contrast, the preoperative severity classification was not an independent determinant after adjustment. Consistent with this finding, stenosis morphology assessed using the Schizas classification was not significantly associated with preoperative balance impairment, suggesting that radiological severity alone may not directly reflect functional balance deficits in patients with LSS.
ROC analysis showed that the MCID for the Brief BESTest total score was consistently 2 points across all patients and within both severity subgroups (Fig. 3). The severe group demonstrated excellent discriminative ability (AUC=0.83), indicating that the Brief BESTest is sensitive for detecting clinically meaningful postoperative changes, even in patients with substantial preoperative impairment. The uniform 2-point threshold across severity levels suggests that this MCID is robust and broadly applicable to the LSS population.
Previous studies have reported larger MCIDs in other clinical populations, such as 3 points after total knee arthroplasty or pulmonary rehabilitation for chronic obstructive pulmonary disease, and 4.5 points in subacute stroke patients [18–20]. These discrepancies likely reflect differences in the underlying pathophysiology and recovery potential; conditions such as stroke or chronic obstructive pulmonary disease often involve widespread multisystem deficits that allow for greater absolute improvements in balance performance. In contrast, patients with LSS exhibit a mixed pattern of impairments in which subsystems that improve readily after surgery, such as anticipatory postural adjustments and gait stability, coexist with subsystems that show inherently limited recovery, such as stability limits. This imbalance may restrict the overall score gains and contribute to the smaller MCID observed.
Although the overall responsiveness of the Brief BESTest was classified as moderate (SRM=0.51), the absolute mean improvement of approximately 2.5 points should be interpreted in the context of the 24-point scale. In older adults with LSS, moderate statistical responsiveness does not necessarily indicate a large perceived clinical improvement. This may be partly explained by the heterogeneous responsiveness of the Brief BESTest subsystems, as some components are more amenable to improvement following spinal surgery, whereas others show limited postoperative changes.
Anticipatory postural adjustments are assessed using the OLS [6,7]. The OLS primarily activates the gluteus medius muscle (GMed) [21], which is frequently affected in patients with L4/5 LSS [22]. In our cohort, 73% of the patients presented with L4/5 stenosis, likely contributing to preoperative GMed weakness and reduced OLS performance. Previous studies have also reported postoperative improvements in GMed strength and reductions in side-to-side asymmetry after LSS surgery [23]. Because OLS performance depends heavily on GMed function, postoperative recovery of GMed strength may have contributed to the improved OLS scores in our cohort.
Gait stability is evaluated using the TUG, one of the most commonly used measures for assessing functional limitations in patients with LSS [6,7,24]. Because the TUG captures multiple components of mobility, including standing up, walking, turning, and dynamic balance, it is highly sensitive to postoperative improvements. Maldaner et al. [25] reported a postoperative SRM of 0.67, which is consistent with the SRM of 0.68 observed in our severe group. This moderate responsiveness may be explained by postoperative reductions in neurogenic claudication, lower extremity pain, and gait instability, all of which collectively enhance dynamic balance and walking capacity.
In contrast, the subsystem with the lowest responsiveness was stability limits, as assessed by the FRT [6,7]. Because FRT performance depends heavily on spinal mobility, it is typically reduced in patients with LSS compared with healthy individuals [26,27], likely contributing to the limited postoperative change. Wada et al. [28] recently reported that delayed lumbar motion during a forward bending–righting task was significantly associated with a fall history in patients with LSS, suggesting that restrictions in lumbar movement range and speed may persist even after decompression surgery. Although stability limits showed limited postoperative changes in our cohort, a recent randomized controlled trial of telerehabilitation after lumbar decompression surgery reported improvements in FRT performance, suggesting that targeted postoperative mobility and balance training may enhance recovery in this subsystem [29].
This study had several limitations. First, it was conducted at a single tertiary care institution, which may limit the generalizability of the findings. Second, the MCID was derived using the GRC, a subjective anchor that may be susceptible to recall bias. Third, the Japanese Orthopaedic Association (JOA) score was not assessed because it was not included in the predefined outcome measures, precluding direct comparison with studies using JOA-based outcomes. Finally, objective evaluations of lower-limb muscle strength, including the GMed, were not performed, limiting the interpretation of the relationship between muscle function and balance recovery.
Conclusions
This study showed that the Brief BESTest has moderate responsiveness to postoperative balance recovery in patients with LSS, with notable improvements in anticipatory postural adjustment and gait stability. The MCID was consistently identified as a 2-point change across all severity levels.
Key Points
The Brief BESTest demonstrated moderate responsiveness to postoperative balance improvements in patients with lumbar spinal stenosis
An minimal clinically important difference of 2 points in the Brief BESTest total score was identified, supported by receiver operating characteristic (ROC) analysis (area under the ROC curve=0.79) and bootstrap validation
Anticipatory postural adjustments and gait stability showed the greatest improvements, whereas stability limits exhibited minimal changes.
Notes
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Funding
This research was supported by the Japan Organization of Occupational Health and Safety (JPJOHAS2025FH25).
Author Contributions
Conceptualization: MT. Data curation: TO, HI, HF, EO, YY, KU. Writing–original draft: TS. Writing–editing draft: SA, TK. Final approval of the manuscript: all authors.
