6-minute walk test and the 3-metre backward walk test are sensitive to postoperative gait improvement in idiopathic normal pressure hydrocephalus

Abstract

Background

Gait impairment is a core symptom in idiopathic normal pressure hydrocephalus (iNPH), but commonly used tests like the 10-metre walk test (10MWT) may underestimate postoperative improvements. We evaluated the 6-minute walk test (6MWT) and 3-metre backward walk test (3MBWT) as outcome measures compared with 10MWT and Timed Up and Go (TUG) and explored sex-related and age-related differences in shunt-operated patients with iNPH.

Methods

A retrospective study of 291 shunted patients with iNPH (62.5% male; mean age 73.3±6.2 years) was conducted. Gait function was assessed preoperatively and at a median of 5.4 months postoperatively. Outcomes were analysed using linear regression, with age and sex as covariates; improvement was described as per cent change from baseline. Correlations between tests were assessed with Spearman coefficients.

Results

Postoperative changes were observed for all gait measures: 3MBWT time (40% reduction (95% CI 36% to 44%)), 6MWT distance (32% increase (95% CI 23% to 43%)), TUG time (28% reduction (95% CI 24% to 32%)) and 10MWT time (19–23% reduction (95% CI 16% to 25%)) at maximal speed and self-selected speed. Women performed worse than men both before and after shunt insertion, but improvement did not differ by sex when adjusting for baseline performance. Each 5-year increase in age reduced gait improvement by 3–8% (p<0.05). Preoperatively, 6MWT correlated strongly with 10MWT (ρ=−0.81) and 3MBWT with TUG (ρ=0.75), while the correlation between 6MWT and 3MBWT was moderate (ρ=−0.58) (all p<0.001).

Conclusions

The 3MBWT and 6MWT capture postoperative gait changes and may provide complementary information to conventional gait assessments in iNPH.

WHAT IS ALREADY KNOWN ON THIS TOPIC.

WHAT THIS STUDY ADDS

• The 6-minute walk test (6MWT) and 3-metre backward walk test (3MBWT) are sensitive to gait improvement after shunt surgery and capture additional dimensions as compared with standard tests, including endurance and balance, which are central to iNPH-related gait disturbances.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Incorporating the 6MWT and 3MBWT may increase sensitivity in detecting shunt-related improvement and better reflect the range of gait and balance dysfunction in iNPH compared with short straight-line walking tests alone.

Introduction

Idiopathic normal pressure hydrocephalus (iNPH) is a common and reversible neurological disease with disturbed gait and balance as the clinical hallmark, in addition to cognitive impairment and urinary incontinence. Shunt surgery improves the majority of patients, predominantly in their gait and balance, but carries some risks.¹ Early intervention, before further disease progression, results in more favourable outcomes, underscoring the importance of timely treatment.² The gait disturbance in iNPH is typically hypokinetic with reduced walking speed, decreased step length, increased double support phase, shuffling and widened step width.^{3 4} Slowed gait during straight walking is only one component of a broader motor impairment,⁵ and cognitive deficits may further affect gait, especially under dual-task conditions.⁶

Sensitive and specific outcome measures targeting iNPH associated gait and balance impairments are essential not only for early detection but also for accurately evaluating post-surgical improvements. To assess baseline impairments and postoperative outcomes in gait and balance function, various measures could be used. In the Japanese guidelines, third version, the Timed Up and Go (TUG) or a short-distance straight walking test such as the 10-metre walk test (10MWT) are recommended.⁷ In a recent Cochrane review of the effect of shunt treatment in iNPH, as well as in a randomised trial of shunt effectiveness, gait speed at the 10MWT was used as the primary outcome measure.^{8 9}

Conventional metrics, such as the 10MWT, which primarily assess gait pace, may not detect subtle deficits in other gait domains, such as dynamic balance,¹⁰ and may be insufficiently sensitive to identify change after cerebrospinal fluid (CSF) diversion.^{11 12} In our clinical experience, additional tests, such as the 6-minute walk test (6MWT) and the 3-metre backward walk test (3MBWT), which assess gait and balance features beyond straight-line speed, add valuable information to the clinical picture.

The 6MWT was introduced by the American Thoracic Society as a submaximal exercise test measuring aerobic capacity and endurance.¹³ The 6MWT measures the distance an individual walks in 6 min and is used to estimate baseline performance and compare changes after treatment. It is used in a wide range of diseases, including neurological disorders such as stroke and Parkinson’s disease.¹⁴ Performance on the 6MWT has also been reported in patients with iNPH in two studies, showing improvement after shunt surgery.^{15 16}

The 3MBWT assesses proprioception, protective reflexes, mobility and balance, while walking backward is important for everyday functional activities.¹⁷ The 3MBWT has been validated in several conditions, including Parkinson’s disease, stroke and multiple sclerosis, and its performance correlates with fall risk.17,20 Despite impaired balance and poor postural control being core features of iNPH,³ backward walking and the 3MBWT specifically has only been reported in two previous studies.^{12 21}

The present study aimed to evaluate the utility of the 6MWT and the 3MBWT as outcome measures in patients with iNPH, in comparison to two established gait tests, 10MWT and TUG. Further we explored age-related and sex-related differences and investigated associations among the tests.

Material and methods

This retrospective study included consecutively enrolled patients diagnosed with iNPH according to international guidelines,³ who underwent shunt surgery at the Hydrocephalus Research Unit in Gothenburg between 1 January 2016 and 31 December 2022. Patients were eligible for inclusion if preoperative and postoperative clinical assessments of 6MWT and Gothenburg iNPH Scale introduced by Hellström²² were available.

Symptoms and signs were scored preoperatively and in median 5.4 months postoperatively on the iNPH Scale,²² with the total score as a sum of four domain scores: gait, balance, cognitive function and continence, where the gait domain is given twice the weight of the other domains. Data from normative studies were used to establish normal score limits, with scores ranging from 0 to 100, where 100 represents the performance of a healthy age-matched normal population.²²

The modified Rankin Scale (mRS; 0–5, higher scores indicating worse function) assessed functional disability and dependence, while the Mini-Mental State Examination (MMSE; 0–30, higher scores indicating better function) assessed global cognitive function.²¹

A standardised clinical examination was conducted by a physiotherapist preoperatively and postoperatively to ensure patient safety. Walking aids were permitted during all tests when needed, and the physiotherapist provided appropriate support and supervision accordingly. The 10MWT was assessed at self-selected speed and at maximal speed, from a stationary start. For the 10MWT self-selected speed, participants were instructed to walk at a comfortable, self-selected pace²²; for the 10MWT- maximal speed, they were instructed to walk as fast as safely possible without running.²³ The tests were conducted twice, with the mean value in seconds (s) calculated for 10MWT self-selected speed and the maximum value (s) registered from the 10MWT maximal speed. Number of steps were registered in each test. In the 3MBWT the patient was instructed to walk backwards quickly, but with a safe pace, for a distance of 3 m, starting from a stationary position. The best time (s) of two trials and corresponding number of steps was registered.²⁴ The 6MWT was performed once in a straight, 30-metre corridor marked with two cones at each end. Participants were instructed to walk back and forth between the cones, covering as much distance as possible during 6 min, without running. The total distance walked was measured in metres. Patients were allowed to pause in an upright position (including leaning towards the wall) if needed. If the patient needed to sit down and rest, the test was ended.¹³

The TUG test was administered with the participant seated in a standard-height chair (45 cm), and the use of armrests was permitted. Participants were instructed to stand up from the chair, walk 3 m at a comfortable, self-selected speed, turn around a floor marker, walk back to the chair and sit down. The time was recorded on the participant’s initial movement from the chair and stopped when seated again with the back against the chair.²⁵ The number of steps taken was recorded, and the best performance from two trials was used in the analysis.

Level of physical activity was assessed with the Saltin-Grimby Physical Activity Level Scale²⁶ by the physiotherapist with the question: ‘How physically active have you been during the last half year/since the surgery? Try to estimate an average’. Leisure-time physical activity was categorised into a four-level scale from (1) physically inactive to (4) regular strenuous physical training for competition sports, several times per week.

Statistical analysis

Descriptive data are presented as counts and percentages for categorical variables, and as mean with SD, median with IQR or geometric mean with coefficient of variation for continuous variables, depending on data distribution.

Statistical analyses were performed using linear regression with preoperative values, postoperative values and postoperative-to-preoperative ratios as outcome measures, and age and/or sex as explanatory variables. All outcome variables were log-transformed prior to analysis due to skewed distributions. Results are presented as fold changes (geometric mean ratios) with 95% CIs, overall and comparing women with men or per 5-year increase in age, calculated by exponentiating the corresponding regression coefficients. Postoperative-to-preoperative ratios were further adjusted for log preoperative values. Fold changes >1 indicate an increase and fold changes <1 indicate a reduction. Percentage change was calculated as 100 × (fold change − 1), with negative values indicating reductions.

Correlations between gait measures were assessed using Spearman correlation coefficients and visualised using heatmaps. Hierarchical clustering was applied to group variables based on the magnitude of their correlations, with dendrograms illustrating relationships using Ward’s minimum variance criterion.²⁷ The strength of correlations was interpreted according to Schober et al,²⁸ where ρ=0.00–0.10 indicates negligible correlation, ρ=0.10–0.39 weak, ρ=0.40–0.69 moderate, ρ=0.70–0.89 strong and ρ=0.90–1.00 very strong. Dropout and subgroup analyses by sex were performed using Mann-Whitney U tests and χ² tests, as appropriate.

Statistical analyses were conducted using IBM SPSS Statistics, V.30.0.0.0 (IBM, Armonk, New York, USA), SAS/STAT Software, V.9.4 (SAS Institute, Cary, North Carolina, USA) and R Software, V.4.4.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Out of 375 eligible participants, 291 fulfilled the inclusion criteria, see figure 1. Patients’ mean age was 75.3 years (SD±6.2) and 182 (62.5%) were males. Demographic and baseline clinical data are shown in table 1.

Figure 1. Flowchart of patient selection and inclusion, showing numbers assessed for eligibility, excluded with reasons and included in the final analysis. iNPH, idiopathic normal pressure hydrocephalus; 6MWT, 6-minute walk test.

Table 1. Baseline demographic and clinical characteristics of the study cohort.

	Preoperative All patients (n=291)	Postoperative All patients (n=291)
Age, mean (SD)	75.3 (6.2)
Male, n (%)	182 (62.5)
Total iNPH score, mean (SD)	54.5 (14.8)	66.7 (17.6)
iNPH scale gait, mean (SD)	47.7 (21.9)	66.0 (25.3)
iNPH scale balance, mean (SD)	65.8 (15.2)	71.6 (14.7)
iNPH scale neuropsychology, mean (SD)	54.7 (18.6)	60.8 (20.0)
iNPH scale continence, mean (SD)	56.8 (26.2)	69.2 (28.1)
Mini-Mental State Examination, mean (SD)	25.5 (1.0)	25.7 (3.6)
Modified Rankin Scale, Md (IQR)	3 (2–3)	2 (2–3)
Education (years), Md (IQR)	11.0 (14.0–8.0)
Living situation
Own home without assistance, n (%)	148 (51.2)	150 (53.0)
Own home with assistance, n (%)	128 (44.3)	116 (41.0)
Service facility/hospital, n (%)	13 (4.5)	17 (6.0)
Symptom duration before diagnosis (months), Md (IQR)	36.0 (60–24)
Time between diagnosis and surgery (months), Md (IQR)	3.6 (5.0–2.3)
Time between surgery and follow-up (months), Md (IQR)	5.4 (7.5–3.8)
Body mass index (kg/m²), mean (SD)	27 (3.9)
Hypertension, n (%)	179 (61.7)
Diabetes, n (%)	75 (25.9)
Cardiac disease, n (%)	82 (28.4)
Any vascular disease, n (%)	206 (71.0)
Walking aids during 6MWT
No walking aids, n (%)	135 (50.0)	135 (51.9)
Cane, n (%)	14 (5.2)	8 (3.1)
Walker, n (%)	121 (44.8)	117 (45.0)
Fall incidents, last 3 months, Md (IQR)	1.0 (0.0–3.0)	0.0 (0.0–0.0)
Physical activity level, Md (IQR)	2.0 (1.0–2.0)	2.0 (2.0–3.0)

Missing observations (preoperative values): n=1 for Mini-Mental State Examination, time between surgery and follow-up, hypertension, diabetes and any vascular disease; n=2 for education, living situation and cardiac disease; n=3 for fall incidents; n=10 for symptom duration before diagnosis; n=11 for physical activity level; n=19 for body mass index; and n=21 for walking aids during the 6MWT.

Missing observations (postoperative values): n=3 for modified Ranking Scale; n=5 for Mini-Mental State Examination; n=6 for fall incidents; n=8 for living situation; n=22 for physical activity level; and n=31 for walking aids during the 6MWT.

iNPH, idiopathic normal pressure hydrocephalus; Md, median; 6MWT, 6-minute walk test.

Drop-out analysis showed that excluded patients (n=84) had more pronounced preoperative symptoms on the iNPH scale (median 47 (IQR 34–61)) compared with included patients (n=291; median 54 (IQR 45–65); p<0.001), and greater degree of disability (mRS median 3 (IQR 2–4) vs 3 (IQR 2–3); p<0.001). No differences were observed in sex or age between groups (p=0.67 and p=0.73). Among excluded patients, reasons for missing data were as follows: n=25 administrative; n=8 missing preoperative 6MWT evaluation; and n=8 missing for unknown reasons.

Postoperative outcomes in gait assessments

Walking endurance and backward walking improved postoperatively, with 6MWT distance increasing by 32% (95% CI 23% to 43%) and 3MBWT time decreasing by 40% (95% CI 36% to 44%). TUG time decreased by 28% (95% CI 24% to 32%), and 10MWT time by 19–23% (95% CI 16% to 25%) at maximal and self-selected speed (table 2).

Table 2. Preoperative and postoperative gait assessments and relative changes following shunt surgery.

Outcome	Preoperative	Postoperative	Fold change (95% CI)	P value
3MBWT time (s)	14.2 (0.73)	8.6 (0.61)	0.60 (0.56 to 0.64)	<0.001
3MBWT steps	17.5 (0.52)	11.8 (0.44)	0.67 (0.64 to 0.71)	<0.001
6MWT metres	201.9 (0.67)	266.1 (0.70)	1.32 (1.23 to 1.41)	<0.001
TUG time (s)	21.4 (0.55)	15.0 (0.50)	0.72 (0.68 to 0.76)	<0.001
TUG steps	21.2 (0.36)	16.8 (0.36)	0.79 (0.76 to 0.82)	<0.001
10MWT self-selected speed time (s)	15.1 (0.41)	11.7 (0.36)	0.77 (0.75 to 0.80)	<0.001
10MWT self-selected speed steps	22.7 (0.31)	18.3 (0.25)	0.81 (0.78 to 0.84)	<0.001
10MWT max speed time (s)	10.1 (0.32)	8.3 (0.33)	0.81 (0.78 to 0.84)	<0.001
10MWT max speed steps	17.7 (0.24)	15.4 (0.21)	0.86 (0.84 to 0.88)	<0.001

Descriptive data are presented as geometric means with coefficients of variation.

Statistical analyses were performed using linear regression applied to log-transformed postoperative-to-preoperative ratios. Results are presented as fold changes (geometric mean ratios) representing change from preoperative to postoperative assessment.

3MBWT, 3-metre backward walk test; 6MWT, 6-minute walk test; 10MWT, 10-metre walk test; TUG, Timed Up and Go.

Onlinesupplemental tables 1 2 present these measures as median (Md) and IQR stratified by age and sex.

Sex differences

No significant sex differences were observed in age, education, MMSE score, symptom duration, time from diagnosis to surgery or time from surgery to follow-up (all p>0.05).

Males, however, had significantly higher total iNPH scores, indicating less pronounced symptoms, than females (median 58 (IQR 49–66) vs 49 (IQR 39–57); p<0.001), while degree of disability did not differ significantly (mRS median 2 (IQR 2–3) vs 3 (IQR 2–3); p=0.38).

Preoperatively, women performed worse than men in all gait tests (table 3), except for 6MWT, where no statistically significant difference was observed. The fold change (geometric mean ratio; women/men) for preoperative 3MBWT time was 1.48 (95% CI 1.27 to 1.72), indicating that women required 48% (95% CI 27% to 72%) longer time to complete the test compared with men (p<0.001). Postoperatively, women performed worse than men in all gait tests except TUG time and steps.

Table 3. Sex differences in gait performance before and after shunt surgery.

Outcome	Female (n=109)	Male (n=182)	Fold change (95% CI)	P value	Adjusted fold change* (95% CI)	P value
Preoperative
3MBWT time (s)	18.2 (0.77)	12.3 (0.65)	1.48 (1.27 to 1.72)	<0.001	–	–
3MBWT steps	20.8 (0.53)	15.8 (0.49)	1.32 (1.17 to 1.48)	<0.001	–	–
6MWT metres	186.9 (0.67)	211.4 (0.67)	0.88 (0.76 to 1.02)	0.10	–	–
TUG time (s)	24.9 (0.58)	19.5 (0.50)	1.28 (1.13 to 1.44)	<0.001	–	–
TUG steps	24.1 (0.37)	19.7 (0.34)	1.23 (1.13 to 1.33)	<0.001	–	–
10MWT self-selected speed time (s)	17.1 (0.42)	14.0 (0.39)	1.22 (1.11 to 1.34)	<0.001	–	–
10MWT self-selected speed steps	25.1 (0.32)	21.4 (0.28)	1.17 (1.09 to 1.26)	<0.001	–	–
10MWT max speed time (s)	11.4 (0.31)	9.4 (0.30)	1.21 (1.11 to 1.30)	<0.001	–	–
10MWT max speed steps	19.2 (0.26)	16.9 (0.22)	1.14 (1.07 to 1.21)	<0.001	–	–
Postoperative
3MBWT time (s)	10.2 (0.65)	7.7 (0.56)	1.31 (1.15 to 1.50)	<0.001	–	–
3MBWT steps	13.2 (0.44)	10.9 (0.43)	1.21 (1.09 to 1.33)	<0.001	–	–
6MWT metres	238.8 (0.81)	284.0 (0.61)	0.84 (0.72 to 0.98)	0.023	–	–
TUG time (s)	16.1 (0.49)	14.5 (0.51)	1.12 (0.99 to 1.26)	0.08	–	–
TUG steps	17.7 (0.34)	16.3 (0.36)	1.08 (1.00 to 1.18)	0.05	–	–
10MWT self-selected speed time (s)	12.5 (0.38)	11.2 (0.34)	1.12 (1.03 to 1.21)	0.009	–	–
10MWT self-selected speed steps	19.5 (0.27)	17.7 (0.24)	1.10 (1.04 to 1.17)	0.001	–	–
10MWT max speed time (s)	9.0 (0.34)	7.9 (0.31)	1.14 (1.05 to 1.23)	0.001	–	–
10MWT max speed steps	16.6 (0.22)	14.8 (0.18)	1.12 (1.07 to 1.18)	<0.001	–	–
Change
3MBWT time (s)	0.55 (0.61)	0.64 (0.56)	0.87 (0.76 to 0.99)	0.035	1.07 (0.95 to 1.20)	0.27
3MBWT steps	0.63 (0.45)	0.70 (0.46)	0.90 (0.81 to 1.00)	0.05	1.06 (0.97 to 1.16)	0.22
6MWT metres	1.28 (0.77)	1.34 (0.57)	0.95 (0.83 to 1.09)	0.48	0.90 (0.79 to 1.02)	0.10
TUG time (s)	0.67 (0.49)	0.74 (0.43)	0.91 (0.81 to 1.02)	0.11	0.99 (0.90 to 1.10)	0.88
TUG steps	0.74 (0.38)	0.83 (0.33)	0.89 (0.82 to 0.97)	0.006	0.98 (0.91 to 1.06)	0.63
10MWT self-selected speed time (s)	0.74 (0.33)	0.80 (0.32)	0.92 (0.85 to 0.99)	0.033	1.00 (0.94 to 1.07)	0.94
10MWT self-selected speed steps	0.78 (0.29)	0.83 (0.27)	0.95 (0.89 to 1.01)	0.10	1.04 (0.98 to 1.10)	0.17
10MWT max speed time (s)	0.78 (0.25)	0.82 (0.28)	0.95 (0.89 to 1.02)	0.18	1.02 (0.96 to 1.09)	0.50
10MWT max speed steps	0.85 (0.24)	0.86 (0.20)	0.98 (0.93 to 1.04)	0.56	1.06 (1.02 to 1.11)	0.007

Descriptive data are presented as geometric means with coefficients of variation.

Statistical analyses were performed using linear regression applied to log-transformed outcomes. Results are presented as fold changes (geometric mean ratios) representing differences between women and men in preoperative values, postoperative values or postoperative-to-preoperative ratios.

^*Adjusted for log-transformed preoperative values. Adjusted analyses for postoperative values and postoperative-to-preoperative ratios are mathematically equivalent; only the latter are shown.

3MBWT, 3-metre backward walk test; 6MWT, 6-minute walk test; 10MWT, 10-metre walk test; TUG, Timed Up and Go.

In unadjusted analyses, women showed greater postoperative improvement than men in 3MBWT, TUG and 10MWT (table 3). However, after adjusting for baseline preoperative performance, these sex differences were no longer statistically significant; only the 10MWT max-speed steps remained significantly different (6% difference; 95% CI 2% to 11%).

Age differences

Preoperatively, older patients showed poorer gait performance in most tests, except for 3MBWT steps and 6MWT. The fold change for preoperative 3MBWT time was 1.07 (95% CI 1.00 to 1.14), indicating that each 5-year increase in age was associated with a 7% longer completion time (95% CI 0% to 14%; p=0.042; see figure 2 and online supplemental table 3). Postoperatively, increased age was associated with worse performance in all tests. For example, in postoperative 6MWT, each 5-year increase in age corresponded to an 8% decrease in distance walked (95% CI 3% to 14%; p=0.005). After adjustment for baseline preoperative values, older patients consistently showed reduced improvement across all gait metrics (online supplemental table 3).

Figure 2. Association between age and gait performance before and after shunt surgery. Results are shown as fold changes (geometric mean ratios) with 95% CIs per 5-year increase in age, derived by exponentiating regression coefficients. Points represent observed values: preoperative (blue circles) and postoperative (red triangles). Solid blue and dashed red lines represent fitted preoperative and postoperative regression lines, respectively, with shaded bands indicating 95% CIs. Panels display results for the 3MBWT (top), 6MWT (second), TUG (third) and 10-metre walk test (bottom; self-selected and maximal speed). Parallel lines suggest that changes between pre-op and post-op values are independent of age. Further details are provided in online supplemental table 3. 3MBWT, 3-metre backward walk test; 6MWT, 6-minute walk test; TUG, Timed Up and Go.

Associations between gait assessments

Preoperatively, 6MWT showed the strongest correlation with 10MWT self-selected speed (ρ=−0.81, p<0.001) and the weakest with 3MBWT (ρ=−0.58, p<0.001). Similarly, 3MBWT correlated most strongly with TUG (ρ=0.75, p<0.001) and the least with 6MWT (figure 3).

Figure 3. Correlations between gait assessments before and after shunt surgery. Heatmaps display Spearman correlation coefficients between gait assessments for (A) preoperative values, (B) postoperative values and (C) postoperative-to-preoperative change. Hierarchical clustering was applied to group variables based on the magnitude of their correlations, with dendrograms shown to the left. Correlation coefficients are displayed within cells and colour-coded according to strength, with lighter colours indicating weaker correlations and darker shades indicating stronger correlations. All correlations were statistically significant (p<0.001). *For visualisation purposes, 6MWT values were multiplied by −1 to harmonise directionality; consequently, correlations involving the 6MWT should be interpreted as negative. 3MBWT, 3-metre backward walk test; 6MWT, 6-minute walk test; TUG, Timed Up and Go.

Postoperatively, scores demonstrated strong to very strong correlations across most gait assessments (ρ=0.77–0.90) except for the correlation between 6MWT and 3MWT which was moderate (ρ=−0.68; all p<0.001). In contrast, the correlations between postoperative change scores were generally weaker (figure 3).

Discussion

We evaluated the utility of the 6MWT and 3MBWT as outcome measures alongside established tests (TUG and 10MWT). Both tests captured postoperative gait change, and among the gait assessments, 6MWT and 3MBWT showed the weakest correlation, suggesting that they assess different aspects of gait and balance function. Sex did not affect postoperative outcomes, but the relative improvement in gait decreased with each 5-year increase in age.

Gait outcomes

The 6MWT and 3MBWT not only capture gait speed but also reflect performance in dynamic balance, endurance and physical capacity, functions that often are impaired in iNPH.¹⁶

To our knowledge, only two previous studies have reported postoperative outcomes on the 6MWT test in iNPH,^{15 16} and our study includes the largest sample to date. 3MBWT outcomes have been published only once, in our earlier smaller study.²¹

In the present study, both tests demonstrated postoperative improvements consistent with previous findings.^{15 16 21} Furthermore, in this cohort, the 6MWT and 3MBWT demonstrated substantial postoperative change, particularly relative to 10MWT, suggesting that they may serve as robust markers of recovery after shunt surgery. Additionally, given their responsiveness, these tests could also be valuable for assessing short-term effects of large volume CSF drains, for example, the CSF tap test for prediction of shunt responsiveness in clinical practice.^{6 12}

The TUG is recommended for assessing tap test response,⁷ and is widely used in Japan, but it may be less sensitive to shunt-related improvements in patients with mild gait impairment.^{15 29} Nikaido et al found that patients with preoperative TUG <20 s improved on the 6MWT but not the TUG.¹⁵ In our cohort, preoperative TUG values were relatively low (median 20.4 (IQR 14.4–29.1) s), yet significant postoperative gains were observed, possibly due to our larger sample. Yamada et al suggest using additional quantitative tests for mild gait disorder (TUG <20 s), as the TUG alone may not capture subtle improvements.²⁹

In the present study, the 6MWT and 3MBWT appeared to be similar or slightly more sensitive to gait improvement than the TUG. Both the 3MBWT and TUG correlate with fall incidence and place greater demands on dynamic balance and postural control than the 10MWT. The TUG assesses functional mobility, dynamic balance and turning ability, whereas the 3MBWT further stresses proprioceptive control and postural adjustments through retrograde gait.³⁰ In contrast, the 6MWT primarily measures aerobic endurance, continuous walking capacity and fatigue resistance over longer distances.³¹

We found only moderately strong correlations between 6MWT and 3MBWT both preoperatively and postoperatively, indicating that these tests assess different aspects of gait and balance and thus provide complementary information. Postoperative changes were even less strongly correlated, suggesting that each test captures distinct domains of functional improvement after shunt surgery in iNPH, supporting their use as outcome measures.

Sex differences

Although women had worse gait performance than men both preoperatively and postoperatively, adjusted analyses showed no substantial sex differences in postoperative improvement, indicating that men and women benefit equally from shunt surgery.

The reason for women’s poorer preoperative gait remains unclear, as age, symptom duration and timing of surgery did not differ. Sex-related physiological differences may also influence test performance, highlighting the value of sex-matched and age-matched normative data for future studies.³² Some previous studies suggest worse gait and function in women with iNPH,^{33 34} while reports on sex disparities in prevalence are conflicting.^{35 36}

Notably, women composed only about one-third of our cohort, despite representing more than half of the elderly population in the region, and exhibited more severe gait impairments than men. This raises questions about whether symptoms in women manifest differently, are overlooked or whether diagnostic criteria, such as radiological markers, are less sensitive^{37 38}; Brix et al suggest that a lower Evans index threshold for women may improve diagnostic accuracy.³⁷ It remains unclear whether observed sex differences reflect true disparities in presentation, disease prevalence or potential gender bias in diagnosis and referral, as seen in other neurological conditions such as stroke.³⁹ These findings highlight the need for further research into sex-specific differences in iNPH care.

Age differences

For most preoperative gait measures, except for 3MBWT steps and 6MWT, higher age was associated with worse walking ability. Regarding shunt outcomes, increasing age was linked to reduced improvements across all tested gait measures. This may be explained by age-related factors such as multisensory decline, a greater prevalence of comorbidities and diminished overall mobility.⁴⁰

Across age groups, 6MWT distances were clearly below systematic-review-derived norms,³² and remained reduced after shunt surgery; for example, patients >80 years improved from median 204 m to 240 m but did not reach the expected 325 m, suggesting further rehabilitative interventions may be beneficial.

Interestingly, some gait outcome measures were less influenced by age (eg, 10MWT self-selected speed (steps) and max speed (steps)). As shown in figure 2, the preoperative and postoperative fold-change lines remain nearly parallel, suggesting that these aspects of gait function are relatively independent of age. However, across the cohort, percentage improvements in 10MWT were less pronounced, whereas TUG and 3MBWT (time) showed high sensitivity to shunt-related improvements while maintaining nearly parallel preoperative and postoperative trajectories (figure 2), making them suitable for older patients.

In contrast, as indicated by figure 2, for younger age groups, 6MWT may be more appropriate. It is highly sensitive to shunt-related improvements and provides valuable insights into the patient’s fitness level. This makes it a useful tool for identifying candidates who could benefit from presurgical rehabilitation programmes to optimise outcomes.

Strengths and limitations

A strength of the current study is that it comprises a large cohort based on real-world data of consecutively selected patients with iNPH diagnosed according to international guidelines. The patient population is representative of the Swedish county of Västra Götaland as the region is served by a single specialised centre that assesses all patients with iNPH listed within the area. The dataset is clinically relevant, reflecting patients undergoing standard diagnostic procedures and shunt surgery according to routine clinical practice. However, the study has some limitations. The dataset was selected based on patients who could complete the 6MWT both preoperatively and postoperatively, which may have resulted in a slightly skewed dataset and potential inclusion bias. On the other hand, as only 31 out of the original 375 population (8.3%) was excluded due to being unable to perform the 6MWT, we believe it shows a high feasibility for the practical application of the test.

A dropout analysis of excluded patients revealed they had more severe baseline impairment on the mRS and iNPH scale compared with those included in the study (n=291), though no significant differences in age or sex were observed. This suggests a potential selection bias toward less severely affected individuals, which may affect the generalisability of the results. Although the analyses accounted for age and sex, the potential influence of the comorbidity burden was not considered, which may have impacted the findings. Another limitation is that age and sex adjusted values for the gait assessments were not used.

Furthermore, the study reflects gait outcomes observed at the first postoperative follow-up, with a median follow-up time of 5.4 months. While these findings provide valuable insights into short-term results, the long-term trajectory remains uncertain.

Conclusions

Assessing gait problems from multiple angles using sensitive tests is essential in iNPH management. In the present study, the 3MBWT, which is quick and easy to use, could reliably detect gait changes irrespective of age and correlated strongly with TUG, while demonstrating high sensitivity to postoperative gait improvement. The 6MWT, which provides information on overall physical fitness that may guide rehabilitation interventions, also showed substantial postoperative improvement. Incorporating these tests as complementary gait measures covering multiple domains of gait and balance function may improve outcome evaluation and iNPH management across different age groups and sexes.

Data availability statement

Data are available upon reasonable request.