Balance Training for People With Multiple Sclerosis Using Nintendo Wii Fit: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Balance Training for People With Multiple Sclerosis Using Nintendo Wii Fit (NCT01299025), a randomized na study. The primary outcome is Change in Timed Up and Go Test From Day 1 to Day 42.

2. Clinical Question

3. Trial Design

The registry describes a randomized, single-masked, parallel-group trial. Participants were allocated to Nintendo Wii Fit training or control. The record does not identify who was masked, describe allocation concealment, or explain how randomization was implemented; these details are not reported in the registry record.

The training program lasted six weeks, with sessions of 30 minutes two times per week. Control participants received no treatment and could be physically active as usual. The registry records an enrollment of 84 and lists the study status as completed. The study start is recorded as 2010-09, and primary completion as 2011-06. The trial phase is listed as NA.

The registry summary says that an estimation of study power indicated that 29 people in each group were necessary to detect a difference in the primary outcome, the Timed Up and Go test. The assumptions underlying that estimate, including the target difference and statistical procedures, are not reported in the registry record. This statement is distinct from the later planning illustrations based on the observed effect.

The posted primary-outcome group summaries include 39 control participants and 41 training participants. The registry also lists four outcome measures and reports three secondary outcomes: the Dynamic Gait Index, Activities-specific Balance Confidence scale, and 12 Item Walking Scale. The brief summary additionally names the Four square step test, Timed stands test, and 25 foot walking test. Results for those measures are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Change in Timed Up and Go Test From Day 1 to Day 42, measured over Measured at day 1 (before the training period) and at day 42 (after the training period).

The primary endpoint is change in the Timed Up and Go test from day 1, before training, to day 42, after the training period. It is measured in seconds. The test records the time taken to rise from a chair, walk 3 meters, turn, walk back, and sit down. It combines standing, walking, turning, and sitting into one timed task.

As a general statistical point, a timed functional test can be analyzed as a continuous outcome: each participant has a numerical result, and group summaries can describe average change and variation. Comparing changes between randomized groups addresses whether the observed pattern over the study period differed by assigned group. The registry presents the group mean changes and standard deviations, as well as a mean difference calculation.

The registry’s primary outcome is relevant to the stated study question because the brief summary identifies the Timed Up and Go test as the primary measure for detecting a difference between groups. However, the record does not provide a threshold for interpreting a change as important to participants, nor does it report a minimum clinically important difference. The numerical results therefore describe measured changes and their statistical uncertainty; they do not by themselves establish the practical importance of a difference.

7. Results Posted in the Registry

• Training With Nintendo Wii Fit: mean -0.80 versus 0.10 on control; difference -0.90 (95% CI -1.89 to 0.09), standardized effect d = -0.40; z-test p = 0.074.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change in Score on the Dynamic Gait Index From Day 1 to Day 42

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Training With Nintendo Wii Fit: mean 1.78 versus 1.00 on control; difference 0.78 (95% CI -0.16 to 1.72), standardized effect d = 0.36; z-test p = 0.105.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Training With Nintendo Wii Fit and the control. The interval gives the range of effects that remain compatible with the data.

Change in the Activities-specific Balance Confidence (ABC) Scalefrom Day 1 to Day 42

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Training With Nintendo Wii Fit: mean 5.00 versus 2.70 on control; difference 2.30 (95% CI -3.46 to 8.06), standardized effect d = 0.17; z-test p = 0.434.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Training With Nintendo Wii Fit and the control. The interval gives the range of effects that remain compatible with the data.

Change in the 12 Item Walking Scale From Day 1 to Day 42

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Training With Nintendo Wii Fit: mean -5.90 versus -3.95 on control; difference -1.95 (95% CI -8.63 to 4.73), standardized effect d = -0.13; z-test p = 0.567.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Training With Nintendo Wii Fit and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover gait and balance confidence: change in the Dynamic Gait Index, the Activities-specific Balance Confidence scale, and the 12 Item Walking Scale. Across these measures, the recalculated mean differences are uncertain, with each reported interval including zero. This overall pattern does not provide a consistent, clearly separated secondary signal in the summary-statistic comparisons.

These secondary outcomes are related to the study’s stated interest in balance capacity and self-efficacy, while the primary endpoint is a timed mobility test. They offer complementary descriptions of function and confidence rather than direct replications of the primary measure. The registry does not report sponsor analyses for these secondary outcomes, so their recalculated comparisons should not be treated as sponsor-confirmed findings.

Unadjusted secondary results carry less weight when several outcomes are examined without a reported multiplicity plan. As a general statistical principle, repeated testing can increase the chance of at least one apparently noteworthy result arising by chance. The registry does not report a testing hierarchy or adjustment, and it does not state how secondary outcomes were selected for analysis. The secondary results are therefore best regarded as exploratory and interpreted alongside the primary result, not as independent confirmation of it.

The brief summary names other measures, including the Four square step test, Timed stands test, and 25 foot walking test, but results for those measures are not reported in the registry record. Their absence means the overall pattern across all measures mentioned in the summary cannot be assessed from the posted results.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change in Score on the Dynamic Gait Index From Day 1 to Day 42
• Change in the Activities-specific Balance Confidence (ABC) Scalefrom Day 1 to Day 42
• Change in the 12 Item Walking Scale From Day 1 to Day 42

9. Statistical Interpretation

10. Statistical Methods Explained

What does the mean difference describe?

The posted primary summaries give a mean change for each group. The recalculated mean difference is the training-group mean minus the control-group mean, yielding -0.9 seconds. This arithmetic comparison summarizes the distance between the two group averages. It does not show how every participant changed, and it does not account for individual-level information beyond what is represented in the posted summary statistics.

As a general explanation, a mean is sensitive to the size of unusually high or low observations. A mean difference is most informative when the group summaries are appropriate for the outcome distribution and the chosen analysis. The registry does not post participant-level data, distribution plots, or diagnostic assessments, so the shape of the outcome distributions cannot be checked from the record.

Why are there two primary p-values?

The sponsor’s reported method is Wilcoxon (Mann-Whitney), a rank-based comparison. In general, this method compares the ordering of observations between groups rather than directly testing a difference in means. Its result can be influenced by differences in distributions and is not simply a test of the posted mean difference. The sponsor reports p <0.01 but does not post a corresponding effect estimate or interval.

The recalculated p-value of 0.074 is an unadjusted comparison derived from the posted group means, standard deviations, and sample sizes. It addresses a mean difference using a different statistical approach. Because the estimand and assumptions are not identical, the two p-values are not interchangeable. The record does not report why the sponsor selected the rank-based method, whether it was prespecified, or whether other analyses were conducted.

What does unadjusted mean?

An unadjusted p-value is reported without an adjustment for multiple comparisons. This matters because the record includes a primary outcome and three posted secondary outcomes, and the brief summary names additional measures. As a general statistical principle, examining more outcomes creates more opportunities for a small p-value to occur by chance when no adjustment or prespecified hierarchy is used. The registry does not report a multiplicity strategy, so the degree of protection against that issue is unknown.

How should the intervals be understood?

As a general explanation, an interval around an estimated difference represents a range of values compatible with the data and the method’s assumptions. A wider interval indicates less precision than a narrower one. For the primary recalculation, the interval crosses zero. The confidence level for the posted intervals is not stated in the registry record, so it should not be assigned a level based on assumption.

The methods used for the recalculated intervals are based on the posted summary statistics and are not the sponsor’s Wilcoxon analysis. They therefore do not supply an interval for the sponsor’s reported rank-based result. The registry reports no interval for the sponsor analysis.

Why can spread matter?

As a general statistical explanation, standard deviations describe how dispersed participant values are around a group mean. Greater variability makes a group average less precise when the sample size is unchanged. The primary standard deviations are 2.1 seconds for control and 2.4 seconds for training. The recalculation uses the posted spreads and sample sizes; it does not adjust for baseline values or other covariates.

Over-dispersion is a concept that arises when variability is greater than a model’s assumptions allow. If a statistical model underestimates variation, it can produce intervals that are too narrow and p-values that are too small. The registry does not report a model-based dispersion assessment for this continuous endpoint, so there is no basis for claiming that over-dispersion occurred. More generally, checks of distributional assumptions and the handling of outliers are not reported in the registry record.

The registry lists the primary outcome as a mean with standard deviation, but the sponsor’s reported rank-based method does not directly test a difference in those means. That distinction is central to understanding why the posted sponsor result and the summary-statistic recalculation can differ without one p-value being a direct correction of the other.

11. Confidence Intervals and Precision

The recalculated primary mean difference is -0.9 seconds, with a reported interval from -1.887 to 0.087. Because the interval includes zero, the summary-statistic calculation is compatible with a range of possible mean differences on both sides of zero. The interval is not especially narrow relative to the point estimate, so the estimate leaves uncertainty about the size and direction of the underlying mean difference.

The confidence level attached to this interval is not reported in the registry record. It should therefore be interpreted as the interval presented with the recalculation, without assuming a particular confidence percentage. It also belongs to the recalculated mean comparison, not to the sponsor’s Wilcoxon analysis. The sponsor’s analysis has no posted effect estimate or interval in the registry record.

As a general statistical explanation, an interval communicates precision more fully than a p-value alone: it shows a range of estimates compatible with the analysis, while a p-value summarizes evidence against a null hypothesis under specific assumptions. Neither statistic describes the probability that the treatment effect is a particular value. Neither resolves limitations in the analysis method or missing information about the study design.

The posted primary group standard deviations, 2.1 seconds and 2.4 seconds, indicate variation among the individual changes. The number of observations contributing to the posted group summaries is 39 for control and 41 for training. The registry does not report why those counts differ from the overall enrollment or whether any participants lacked an endpoint measurement. That information would help readers assess the precision and representativeness of the estimates.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size illustrations are planning calculations based on fractions of the observed effect, not the trial’s original design calculation. They assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For these continuous outcomes, the listed values are participants per arm, not a total across both arms and not total events. The event-count convention applies to time-to-event outcomes; none of the posted outcomes is presented as a time-to-event measure.

The 80% power column gives the planned participants per arm for a hypothetical study with an 80% chance of detecting the specified effect under the stated assumptions, if that effect and assumptions were correct. The 90% power column describes the corresponding planning target for a 90% chance. These are prospective planning illustrations, not observed power.

Rows describe planning for 100%, 75%, and 50% of the observed effect. The smaller the effect a future study aims to detect, the more participants it generally needs, all else equal. The figures are conditional on the calculation’s assumptions and should not be read as a recommendation or as evidence that the observed effect is stable. The registry’s own brief summary separately reports an estimate of 29 people in each group as necessary to detect a difference in the primary outcome. The assumptions behind that original estimate are not reported in the registry record, so it cannot be directly compared with these illustrations as if both used the same design choices.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.900; pooled standard deviation σ = 2.259.
2. n = 2 × (1.96 + 0.842)² × 2.259² / -0.900² = 98.9, rounded up to 99 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.9                      # observed difference in means
sigma <- 2.25882883200735                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1299025)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 99 (80% power), 133 (90% power)
 75% of observed effect: n per arm = 176 (80% power), 236 (90% power)
 50% of observed effect: n per arm = 396 (80% power), 530 (90% power)
power.t.test: 99.9 
Simulated power at n = 99 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports that the sponsor used a Wilcoxon (Mann-Whitney) analysis for the primary comparison and obtained p <0.01. No sponsor effect estimate or interval is posted. Separately, a recalculation from the group means, standard deviations, and sample sizes gives a mean difference of -0.9 seconds, an interval from -1.887 to 0.087, and an unadjusted p-value of 0.074.

These results differ because the recalculation compares means using summary statistics, while the sponsor’s rank-based method compares the ordering of observations. They are not two versions of the same analysis, and the recalculation cannot replace the sponsor’s own result. Given the recorded discrepancy, the sponsor’s own analysis takes precedence as the reported trial analysis. The summary-statistic recalculation is useful as a separate perspective on the posted means, but it should not be used to overwrite the registry’s reported test.

The registry does not explain the reason for choosing the Wilcoxon method, identify whether it was planned in advance, or provide enough details to reproduce the sponsor’s analysis. It also does not post a rank-based effect estimate or interval, which limits assessment of the magnitude and precision of that result.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary endpoint and three secondary outcomes, while the brief summary also names additional measures. It does not report whether statistical testing followed a prespecified hierarchy, whether any adjustment was made for multiple outcomes, or how the primary and secondary hypotheses were distinguished. This matters because the more comparisons are explored, the greater the opportunity for chance findings. The sponsor reports p <0.01 for the primary analysis, but the record does not state whether that value is adjusted.

Information about missing data is also limited. The registry lists enrollment of 84, while the posted primary summaries include 39 participants in control and 41 in training. The reason for the difference in counts is not reported in the registry record. The record does not explain whether participants withdrew, lacked measurements, were excluded from analysis, or were handled through an imputation or other missing-data method. No conclusion about the reasons or consequences of the differing counts can be drawn from the record alone.

The analysis population is not specified. It is not reported whether the analysis included all randomized participants according to their assigned group, only participants with observed outcome measurements, or another population. The record also does not report adherence to training, deviations from assigned treatment, or how such events were handled analytically.

The single-masked designation does not identify which parties were masked. Allocation concealment, prespecification of the sponsor’s analysis method, and the handling of protocol deviations are not reported in the registry record. These omissions do not establish that the methods were inadequate; they limit what can be verified about the analysis and how securely the results can be interpreted.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The safety record lists 42 participants at risk in each group for serious adverse events and other adverse events. It records zero affected participants in both categories for the control group and for the training group. These are the registry’s posted counts; the record does not provide event descriptions because no affected participants are listed.

Deaths are not reported in the registry record. The safety denominators differ from the participant counts in the posted primary-outcome summaries, and the reason for that difference is not reported. The safety figures should therefore be read as the registry’s adverse-event accounting, not assumed to use the same analysis population as the outcome summaries.

As a general statistical point, a recorded zero does not establish that an event is impossible or that a risk is absent. Safety comparisons can be imprecise when event counts are low, and the record does not provide a detailed safety analysis or an interval for these counts. These figures describe what the registry records, not a broader assessment of safety.

17. What the Result Does, and Does Not, Mean

The results describe changes measured from day 1 to day 42 in a randomized comparison of Nintendo Wii Fit training with a control condition. The primary summary-statistic recalculation estimates a mean difference of -0.9 seconds in the Timed Up and Go change, with an interval that includes zero. The sponsor’s reported Wilcoxon analysis gives p <0.01, but the registry does not post its effect estimate or interval. These findings must be kept distinct because they come from different methods.

The posted numbers do not, by themselves, establish the practical importance of a change. The registry does not provide a threshold for a meaningful difference, nor does it report longer-term follow-up. The measurements concern the stated study period and cannot establish whether any differences persisted beyond day 42.

The results also do not resolve uncertainty about the analysis population, missing observations, or multiplicity. The difference between enrollment and the primary-outcome group counts is not explained. The registry does not report whether the sponsor analysis was prespecified or whether secondary comparisons were adjusted. These gaps matter when judging how robustly the reported findings represent the randomized groups and how much weight to place on individual p-values.

As a general interpretation principle, statistical evidence, effect magnitude, precision, and study methods are separate considerations. A p-value does not measure the size or importance of an effect, and an interval does not remove uncertainty about study conduct or analysis choices. The registry supports a description of the recorded results and methods, but not a definitive claim about the broader value of the training intervention.

18. Why This Trial Matters Statistically

This trial is statistically instructive because it presents a sponsor-reported rank-based test alongside group means and standard deviations that permit a separate mean-difference recalculation. The two analyses yield different p-values, illustrating why a reported test statistic must be interpreted in light of the method and the question it addresses. A rank-based comparison and a comparison of means are not interchangeable merely because they use the same groups and endpoint.

The record also shows why estimates and intervals are important alongside significance tests. The recalculated mean difference is -0.9 seconds, and its interval includes zero. That combination communicates both the estimated contrast and the uncertainty around it. The sponsor analysis, by contrast, is reported only as p <0.01, without a corresponding effect estimate or interval, leaving the magnitude and precision of that analysis less visible.

Other useful statistical questions concern the count of participants contributing to outcome summaries, the absence of a stated missing-data method, and the lack of a reported multiplicity plan. These are not technical footnotes: they affect how readers assess the analysis population, the possibility of selective emphasis, and the reliability of individual comparisons.

Finally, the posted planning illustrations make clear that required sample size depends on the effect a future study aims to detect. Smaller target effects require more participants under the stated assumptions. Those calculations are planning aids based on the observed effect, not a retrospective measure of whether this trial had adequate power.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several trial-specific details are absent from the registry record. It does not identify who was masked, describe allocation concealment, explain the randomization procedure, or state whether the sponsor’s Wilcoxon analysis was prespecified. It also provides no sponsor effect estimate or interval for the primary analysis, making its magnitude and precision difficult to assess from the reported test alone.

The record does not explain why the primary-outcome summaries contain 39 control participants and 41 training participants despite an enrollment of 84. It does not report the analysis population, losses to follow-up, missing-data handling, treatment adherence, or protocol deviations. These omissions restrict assessment of how the analyzed participants relate to those randomized.

The registry does not report a multiplicity strategy, and results for measures named in the brief summary, including the Four square step test, Timed stands test, and 25 foot walking test, are not reported. Confidence levels for the recalculated intervals are not stated. The posted primary outcome is measured through day 42, and longer-term results are not reported in the registry record.

These limitations describe what cannot be verified from the registry record. They do not establish that unreported procedures were not performed, nor do they determine the direction of any resulting bias.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Balance Training for People With Multiple Sclerosis Using Nintendo Wii Fit — NCT01299025.
• Registry record last updated 2015-04-24; status last verified 2015-04.
• Results first posted 2015-04-24.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry describes a completed randomized, single-masked, parallel-group trial of Nintendo Wii Fit balance training for people with multiple sclerosis. Training lasted six weeks, with sessions of 30 minutes two times per week; the control group received no treatment and could remain physically active as usual. Enrollment is listed as 84.

The primary outcome was change in the Timed Up and Go test from day 1 to day 42. The sponsor reports a Wilcoxon (Mann-Whitney) analysis with p <0.01. A separate recalculation from posted group summaries gives a mean difference of -0.9 seconds, with an interval crossing zero and an unadjusted p-value of 0.074. The discrepancy should remain explicit: the sponsor’s analysis takes precedence as the trial’s reported analysis, while the recalculation describes the posted means.

Three secondary outcomes have posted summaries, but no sponsor analyses are reported for them. The registry does not explain the difference between enrollment and primary-outcome summary counts, or report missing-data and multiplicity methods. Safety listings record zero serious and other adverse events among the participants listed at risk in each group; deaths are not reported in the registry record.