Memantine for Executive Dysfunction in Adults With ADHD: A Pilot Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Memantine for Executive Dysfunction in Adults With ADHD: A Pilot Study (NCT01533493), a randomized na study. The primary outcome is Percent Change in Global Executive Composite T-Score on the Behavior Rating Inventory of Executive Function-Adult (BRIEF-A).

2. Clinical Question

3. Trial Design

The registry describes a randomized, double-blind, parallel-group pilot study. It lists the trial status as completed, enrollment as 33, and phase as NA. The registry start and primary-completion dates are 2012-05 and 2013-07. The study was sponsored by Massachusetts General Hospital.

After screening, participants were assigned in double-blind fashion to memantine or a matching placebo for 12 weeks. The brief study description says there was an equal chance of receiving memantine or placebo. Both groups also received open-label OROS-methylphenidate. Thus, the randomized comparison is between two treatment strategies that share OROS-methylphenidate and differ in whether memantine or matching placebo was added.

The posted primary outcome summarizes percent change from baseline to 12 weeks in the BRIEF-A Global Executive Composite T-score. The reported outcome summaries include 12 participants in the memantine group and 14 in the placebo group. The registry does not provide a participant-flow account explaining the relationship between enrollment and the numbers represented in these outcome summaries. Reasons for any missing outcome data are not reported in the registry record.

The registry lists one outcome measure and no posted analyses. It does not report the original statistical analysis plan, a sample-size justification, interim analyses, or detailed procedures for handling missing observations. Those omissions mean the posted summaries can describe the observed groups, but they do not establish how the trial’s investigators planned every aspect of the analysis.

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.

Before assignment: Of the 33 participants enrolled, 26 were randomized to receive the treatment. The remaining 7 participants were either lost to follow-up after signing consent (n=3) or withdrew consent before being randomized (n=4)

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 Percent Change in Global Executive Composite T-Score on the Behavior Rating Inventory of Executive Function-Adult (BRIEF-A), measured over baseline, 12 weeks.

The primary endpoint is percent change from baseline in the Global Executive Composite (GEC) T-score on the Behavior Rating Inventory of Executive Function-Adult (BRIEF-A), assessed at 12 weeks. The BRIEF-A is described as a 75-item checklist covering nine subscales: Inhibit, Shift, Emotional Control, Self-Monitor, Initiate, Working Memory, Plan/Organize, Task Monitor, and Organization of Materials. Items are rated Never, Sometimes, or Often, and the GEC is calculated by totaling the scale items.

The registry says GEC T-scores range from 34-108, with higher scores indicating more executive-function difficulties. In general, a change in a standardized score can provide a structured way to compare participants’ reported difficulties over time. Percent change expresses the outcome relative to its baseline value, although the registry does not explain the precise calculation used for the posted percent-change summaries.

For this study, negative mean changes indicate lower scores at follow-up relative to baseline under the reported percent-change measure. Because higher GEC T-scores indicate more difficulties, a lower score represents a direction of change toward fewer reported difficulties on the scale. That score direction alone does not establish the clinical importance of a difference. The registry record does not report a threshold for interpreting a change as clinically meaningful.

7. Results Posted in the Registry

• Memantine: mean -24.00 versus -21.00 on control; difference -3.00 (95% CI -3.12 to -2.88), standardized effect d = -19.99; z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it is a numerical percent-change measure rather than an event count. The registry specifies that the posted group summaries use means and standard deviations. In general, a mean describes the center of a group’s values, while a standard deviation describes how dispersed those values are around the mean. A mean difference places the two group means on the same scale and preserves the direction of the comparison.

What does the mean difference represent?

The comparison is stated as memantine minus placebo. The group means are -24.0 and -21.0, and the unadjusted recalculated mean difference is -3.0. Because the endpoint is percent change, the difference is described in percentage points, not as a relative percentage difference between the groups. The negative sign indicates that the memantine group’s mean change was lower on this measure than the placebo group’s mean change. This is a description of the direction and size of the observed contrast, not an independent judgment about its clinical meaning.

What does an unadjusted calculation assume?

An unadjusted comparison uses the posted group summaries without accounting for other participant characteristics or covariates. In general, such a calculation is simpler to interpret, but it may not reflect the analysis chosen by trial investigators. Randomization supports the intended comparison, but the registry record does not report baseline balance, the analysis population, or a sponsor model. The posted calculation therefore should be distinguished from a documented prespecified analysis. The registry lists zero posted analyses, and the sponsor’s statistical methods are not reported in the registry record.

How should the p-value be read?

The posted unadjusted p-value is p < 0.001. In general, a p-value compares the observed data with a null hypothesis using a statistical model. It indicates how compatible the data are with that null model, assuming the model and analysis conditions are appropriate. It is not the probability that the null is true, the probability that the result is due to chance, or a measure of clinical importance. The registry record does not describe the p-value’s original analysis procedure; the displayed comparison is an unadjusted recalculation from the posted summaries.

Small samples can make estimates and inferential calculations sensitive to individual observations and to modeling assumptions. The group outcome summaries contain 12 participants for memantine and 14 for placebo. In general, standard errors quantify uncertainty in an estimate and tend to be larger when information is limited or observations are variable. The registry includes standard errors of 0.049 for memantine and 0.035 for placebo, but does not report the underlying participant-level observations or a detailed method for obtaining the original outcome summaries.

Why can variability affect the comparison?

In general, dispersion matters because the uncertainty around a mean difference depends not only on the distance between the means but also on the variability and amount of information in each group. A larger difference relative to the spread can produce a more extreme test statistic, while high variability can make a difference harder to estimate precisely. The posted standard deviations are 0.169 for memantine and 0.132 for placebo. Their scale and calculation are not explained further in the registry record, so they should be read as the posted dispersion values rather than independently reinterpreted.

Over-dispersion is a general concern when observed variation exceeds what a statistical model assumes. If a model understates variability, its standard errors may be too small and intervals too narrow. The registry does not report a model-based analysis or an assessment of over-dispersion for this continuous outcome. The concept is therefore relevant as a general statistical caution, not as a finding about this trial.

What role does the confidence interval play?

A confidence interval gives a range of estimates compatible with the data under the calculation’s assumptions and conveys precision alongside the point estimate. The posted unadjusted recalculated interval for the mean difference is -3.118 to -2.882. It is very narrow relative to the contrast and does not include zero. These properties describe the interval produced from the posted summaries; they do not remove concerns about the limited outcome-summary groups, unreported analysis choices, or the suitability of the calculation. A confidence interval is not a range that contains the true effect with a known probability after the data have been observed.

Overall, the reported values support a clear description of the posted contrast, but the registry’s limited analysis documentation constrains how far it can be interpreted. The recalculated inferential results are not a substitute for a protocol, prespecified analysis plan, or full report of participant-level analysis methods.

10. Confidence Intervals and Precision

The unadjusted recalculated mean difference is -3.0 percentage points, and its posted recalculated 95% confidence interval is -3.118 to -2.882. A confidence interval is a way to express the precision of an estimate under the assumptions used to calculate it. In general, a narrower interval indicates less estimated sampling uncertainty under those assumptions; it does not guarantee that the assumptions are correct or that the estimate applies beyond the study.

This interval does not include zero, and the posted unadjusted p-value is p < 0.001. These inferential summaries are internally aligned in indicating a contrast away from zero under the unadjusted calculation. They do not establish clinical importance. The registry does not report a threshold for a meaningful difference on this percent-change endpoint, and the clinical interpretation cannot be settled by statistical significance alone.

The interval is a recalculation from posted group summaries rather than a sponsor-reported interval documented with an analysis method. Its precision should therefore be read with caution, particularly because the outcome summaries include 12 participants in the memantine group and 14 in the placebo group, while the registry does not report participant-level data, missing-data handling, or the original analysis population. A very narrow interval can reflect the inputs and assumptions of a calculation without representing every source of uncertainty in a trial.

The interval also addresses only the estimated mean contrast for this endpoint and comparison. It does not describe variation in individual outcomes, effects on other measures, safety uncertainty, or how the result might differ in other populations or settings. The registry record does not report secondary posted outcomes or additional analyses that would broaden those conclusions.

11. 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 planning figures are participants per arm, not total events. They use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The assumed effect to detect is the stated fraction of the observed effect, with rows for 100%, 75%, and 50% of the observed effect. The registry record does not report the trial’s original sample-size assumptions or justification; these figures are a planning illustration based on the observed contrast, not a record of the design process.

The 80% power column represents the number of participants per arm that the illustration estimates would be needed for an 80% chance of detecting the assumed effect under those assumptions. The 90% power column gives the corresponding planning figure for a 90% chance. In general, higher target power or a smaller effect to detect requires more participants, all else being equal.

The posted figures are 1 participant per arm for each displayed effect fraction in both power columns. They should not be read as a recommendation that a future study enroll only that number. The calculations are driven by the observed effect and the posted variability, and an unusually large observed effect in a small pilot can produce unstable planning estimates. The normal approximation and equal-allocation assumptions also simplify real design decisions. These values are not observed power; they are hypothetical planning calculations and do not describe the probability that this completed study detected an effect.

12. 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 = -3.000; pooled standard deviation σ = 0.150.
2. n = 2 × (1.96 + 0.842)² × 0.150² / -3.000² = 0.0, rounded up to 1 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 <- -3                      # observed difference in means
sigma <- 0.150094831134631                     # 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(1533493)
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 = 1 (80% power), 1 (90% power)
 75% of observed effect: n per arm = 1 (80% power), 1 (90% power)
 50% of observed effect: n per arm = 1 (80% power), 1 (90% power)
power.t.test: 1.5 
Simulated power at n = 1 per arm: NA (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and zero secondary outcomes posted, along with zero analyses. The single listed primary measure reduces the visible scope for multiplicity across posted outcome measures, but it does not establish that no other outcomes were collected or analyzed. The registry record does not report the protocol’s full outcome hierarchy, the prespecified analysis plan, or whether multiple time points or alternative analyses were considered.

Multiplicity arises in general when several hypotheses are tested and the chance of at least one false-positive result can increase if each is assessed without accounting for the others. Whether and how multiplicity was handled in this study is not reported in the registry record. The posted p-value is unadjusted; no multiplicity adjustment is described. It should therefore be interpreted as the result for this recalculated comparison, not as an adjustment across a documented family of analyses.

The registry gives outcome summaries for 12 participants in the memantine group and 14 in the placebo group, but does not report the number with missing primary outcome data, reasons for missingness, or the method used to address missing values. Missingness can matter because a complete-case summary may differ from an analysis that includes all randomized participants under a stated model. The registry record does not identify whether the posted results follow an intention-to-treat, per-protocol, or other analysis population.

There is also no reported participant-flow account to explain how the outcome-summary counts relate to the enrollment of 33. Baseline characteristics by randomized group, deviations from assigned treatment, and discontinuations are not reported in the registry record. These details would help readers assess the scope and representativeness of the posted estimates, but their absence does not by itself show how any missing data affected the numerical contrast.

14. 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 5% reporting threshold.

Most frequent non-serious adverse event terms

The serious adverse event figures are reported as zero in both randomized groups: 0 among the 12 participants at risk in the memantine group and 0 among the 14 participants at risk in the placebo group. The listed top serious adverse event terms are empty. Death counts are not reported in the registry record.

These figures describe the serious-event counts recorded in the registry, not all possible harms or all symptoms experienced during the study. A zero count in a group does not establish that an event is impossible, nor does it rule out uncommon events. The registry does not provide event definitions, observation-time details, or further information about event collection in this record.

The non-serious adverse event figures are separate from the serious-event figures: the registry records 10 affected participants in the memantine group and 13 in the placebo group. Those counts should not be conflated with serious adverse events. The registry’s serious-event relatedness counts are zero for both groups; interpretation of the safety record remains limited by the small outcome groups and the absence of event-level detail.

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

The posted result describes a randomized comparison of memantine and matching placebo, both given alongside open-label OROS-methylphenidate, on percent change in BRIEF-A GEC T-score at 12 weeks. The group means are -24.0 and -21.0, respectively, and the unadjusted recalculated mean difference is -3.0 percentage points. Because lower GEC T-scores indicate fewer executive-function difficulties, the direction of the mean contrast is toward a lower score in the memantine group. The interval and p-value summarize statistical uncertainty under the recalculation’s assumptions.

This does not establish that memantine caused the entire observed difference, that the difference is clinically important, or that a similar result would occur in other groups or settings. The posted comparison is an unadjusted calculation from summary statistics, and the sponsor’s detailed analysis method is not reported in the registry record. The endpoint is a scale-based measure, and the registry does not report a threshold for a meaningful change.

The study also does not answer whether OROS-methylphenidate alone changes the endpoint, because both randomized groups received it. Nor does the posted comparison address treatment without OROS-methylphenidate or outcomes beyond the registered primary measure. Follow-up for this endpoint is reported at 12 weeks; longer-term outcomes are not reported in the registry record.

The safety counts show no recorded serious adverse events in either group in the registry entry, but they do not demonstrate the absence of risk. The record’s missing details on participant flow, missing outcomes, analysis populations, and adverse-event ascertainment limit broader conclusions. The result is best understood as a pilot-study estimate for a specific randomized add-on comparison and a specific executive-function measure, rather than a definitive conclusion about treatment effects.

16. Why This Trial Matters Statistically

This trial raises several useful statistical questions in a compact example. Its randomized, double-blind, parallel design creates a direct comparison between memantine and matching placebo, while the common open-label OROS-methylphenidate treatment defines the context of that comparison. This makes clear why an effect estimate must be tied to the actual contrast: the posted result concerns the difference between add-on groups, not memantine versus no treatment.

The result also illustrates why an estimate, interval, and p-value answer different questions. The mean difference communicates the direction and magnitude of the observed group contrast, the interval conveys precision under calculation assumptions, and the p-value describes compatibility with a null model. None by itself establishes clinical importance. The registry does not report a clinically meaningful threshold for the BRIEF-A measure.

A further lesson is the sensitivity of planning calculations based on an observed pilot effect. The posted planning values are 1 participant per arm across the displayed effect fractions and power targets. Those figures are not a design recommendation; they demonstrate how observed estimates and simplified assumptions can drive a planning illustration. The original design justification is not reported in the registry record.

Finally, the record shows how limited analysis documentation affects interpretation even when an outcome contrast and inferential summaries are available. No analyses are posted, and missing-data methods, analysis populations, multiplicity procedures, and the original statistical plan are not reported. Distinguishing what is directly recorded from what can only be evaluated with fuller methods and participant-level information is central to sound statistical reading.

17. 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.

Interpretation is limited by the pilot scale of the study and by the outcome-summary groups, which include 12 participants for memantine and 14 for placebo. The registry lists enrollment of 33 but does not provide a participant-flow account linking enrollment to the posted outcome summaries. It is therefore not possible from this record to determine reasons for the difference in counts or how missing primary outcome data were handled.

The registry reports no analyses and does not provide the original statistical analysis plan, model specification, or analysis population. The recalculated mean difference, interval, and p-value are unadjusted summaries and should not be represented as the sponsor’s documented primary analysis. Baseline group characteristics and adherence or discontinuation information are not reported in the registry record.

The endpoint is a percent-change measure, but the exact calculation and any interpretation threshold are not described. The registry does not report secondary outcomes or longer-term follow-up results. The confidence interval and planning illustration depend on posted summaries and their assumptions; neither can capture uncertainty arising from unreported design or analysis choices.

Safety interpretation is also limited: the record gives serious-event counts but no event-level descriptions or detailed ascertainment information, and death counts are not reported in the registry record. These limitations constrain what can be concluded from the registry entry without showing that the posted numerical results are incorrect.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Memantine for Executive Dysfunction in Adults With ADHD: A Pilot Study — NCT01533493.
• Registry record last updated 2014-04-08; status last verified 2014-03.
• Results first posted 2014-02-21.
• 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.

21. Record Summary

NCT01533493 is a completed randomized, double-blind, parallel-group pilot study sponsored by Massachusetts General Hospital. It evaluated memantine hydrochloride versus matching placebo, with open-label OROS-methylphenidate in both groups, over 12 weeks in adults with ADHD and executive function deficits. Enrollment was 33; the registry lists the start as 2012-05 and primary completion as 2013-07.

The primary outcome was percent change from baseline in BRIEF-A Global Executive Composite T-score at 12 weeks. The posted means were -24.0 for memantine and -21.0 for placebo, with an unadjusted recalculated mean difference of -3.0 percentage points and p < 0.001. The registry does not report the sponsor’s analysis methods, missing-data handling, or original sample-size justification.