Roflumilast and Donepezil to Reverse Scopolamine Induced Cognitive Deficits in Healthy Adults: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Roflumilast and Donepezil to Reverse Scopolamine Induced Cognitive Deficits in Healthy Adults (NCT02051335), a randomized phase 1 study. The primary outcome is Change From Baseline in the Verbal Recall Memory (VRM) Total Number of Correct Responses for Delayed Recall at 1 Hour After Scopolamine Administration.

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, placebo-controlled, quadruple-masked, 4-period crossover study in healthy adults. It is listed as Phase 1 and completed. The registry records enrollment of 27, with a start date of 2014-01 and primary completion in 2014-05. The official title describes single oral administrations of roflumilast, alone or in combination with donepezil, to evaluate effects on scopolamine-induced deficits in psychomotor and cognitive function. Scopolamine was also administered as part of the study conditions.

The four treatment conditions were A: Placebo, B: Donepezil 10 mg, C: Roflumilast Dose A, and D: Roflumilast Dose A + Donepezil 10 mg. The registry describes four treatment sequences by letter: ABDC, BCAD, CDBA, and DACB. The posted arm descriptions identify oral tablets and overencapsulated tablets for the study drugs, and describe scopolamine 0.5 mg as a subcutaneous injection in the recorded sequence descriptions. Some sequence text is truncated in the registry record, so full details of each period’s administration are not reported in the registry record.

A crossover design means that treatment conditions are studied across periods within a sequence, rather than assigning every participant to just one parallel treatment group. The posted outcome counts differ by condition: 23 for placebo, 23 for donepezil, 25 for roflumilast, and 27 for the combination. These are the analysis counts shown for the primary outcome, not a basis for adding participants across conditions; the same study is described as a crossover. The registry identifies the primary outcome population as Full Analysis Set participants with data available for analysis at the specified time-point. Details of how randomization was generated, how treatment periods were separated, or how period and carryover effects were handled 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.

Before assignment: Participants with a diagnosis of Alzeimer's disease were enrolled equally in a 4 period, 4 treatment crossover study. Treatment A: Placebo, Treatment B: Donepezil, Treatment C: Roflumilast and Treatment D: Roflumilast + Donepezil. All participants received scopolamine on Day 1 of each treatment period.

Period 1

Period 2

Period 3

Period 4

Analysis population for the primary outcome, as defined in the registry: Participants from the Full Analysis Set with data available for analysis at the given time-point.

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 From Baseline in the Verbal Recall Memory (VRM) Total Number of Correct Responses for Delayed Recall at 1 Hour After Scopolamine Administration, measured over Baseline and 1 hour after scopolamine administration on Day 1 of each treatment period. Baseline is defined as the assessment 1 hour before roflumilast/donepezil administration (3 hours before scopolamine administration)..

The primary endpoint is change from baseline in the Verbal Recall Memory (VRM) total number of correct responses for delayed recall at 1 hour after scopolamine administration. The registry defines baseline as the assessment 1 hour before roflumilast or donepezil administration, which is 3 hours before scopolamine administration. The outcome timeframe is baseline and 1 hour after scopolamine administration on Day 1 of each treatment period.

VRM assesses the ability to encode and later retrieve verbal information. Participants see a series of 18 words, complete immediate recall tasks, and then complete a delayed recall task after approximately 20-30 minutes. The possible score ranges from 0, worst, to 18, best; a higher number of correct responses indicates a better outcome. A negative change from baseline therefore indicates worsening of the score. The endpoint is continuous and is reported as a mean with standard deviation.

This measure fits the study’s stated question because the study aimed to examine cognitive performance after scopolamine-induced impairment. Delayed recall captures retrieval of verbal material after a delay, rather than immediate recall alone. The posted endpoint is a change from a pre-treatment baseline, which describes how the score shifted within each treatment condition. The mean changes are negative in all conditions, but the registry results do not by themselves explain why individual scores changed or establish the clinical importance of the observed differences. A threshold for a meaningful difference is not reported in the registry record.

7. Results Posted in the Registry

• B: Donepezil 10 mg: mean -7.00 versus -6.70 on control; difference -0.30 (95% CI -2.14 to 1.54), standardized effect d = -0.09; z-test p = 0.749.
• C: Roflumilast Dose A: mean -7.40 versus -6.70 on control; difference -0.70 (95% CI -2.58 to 1.18), standardized effect d = -0.21; z-test p = 0.466.
• D: Roflumilast Dose A + Donepezil 10 mg: mean -6.60 versus -6.70 on control; difference 0.10 (95% CI -1.50 to 1.70), standardized effect d = 0.04; z-test p = 0.902.

8. Statistical Interpretation

9. Statistical Methods Explained

A mean change summarizes the average difference between each participant’s baseline score and the score at the specified follow-up assessment. In this trial, the endpoint is the number of correct delayed verbal recall responses, and negative change denotes a worsening of the score. The posted group means describe average changes for each treatment condition. They do not show the distribution of individual responses or whether every participant changed in the same direction.

What does a mean difference represent?

A mean difference compares the average change in one condition with the average change in another. For example, the registry’s reported estimate for donepezil versus placebo is a least-squares mean difference of -0.31. The estimate’s sign indicates the direction of the contrast as reported, while its magnitude expresses the difference in the outcome’s units: correct responses. A value near zero indicates that the estimated average changes are close, but it does not show that the conditions are identical. The sponsor’s estimates are least-squares mean differences from ANOVA, whereas the recalculated estimates are unadjusted differences from posted summary statistics. Those approaches can produce different estimates and uncertainty because they do not necessarily use the same model information.

What does the confidence interval describe?

A confidence interval expresses statistical uncertainty around an estimated difference. In general, a wider interval indicates less precision, while a narrower interval indicates greater precision, conditional on the analysis assumptions. The sponsor reports 95% confidence intervals for the mean differences. Each reported interval for the primary comparisons against placebo extends across zero. In general, an interval that includes zero means the data do not clearly distinguish the estimated contrast from no average difference at the stated confidence level. It does not prove that there is no effect, and it does not assign a probability to the true effect being inside the interval.

The recalculated intervals are also unadjusted estimates based on group means, standard deviations, and analysis counts. They should be read as a separate, simple statistical reconstruction, not as a replacement for the sponsor’s ANOVA. Because this study is a crossover, observations from different treatment periods may be related within a participant. Aggregate group summaries do not preserve the within-participant relationships needed to reconstruct a paired analysis. The recalculations therefore should not be treated as though they reproduce the sponsor’s crossover model.

How should the p-values be read?

A p-value measures how surprising data at least as incompatible with a specified null model would be if that model and its assumptions were correct. It is not the probability that a treatment has no effect, and it does not measure the size or importance of a difference. The sponsor’s p-values for the comparisons against placebo are 0.573, 0.210, and 0.821. The recalculated unadjusted p-values are 0.749, 0.466, and 0.902. These values do not give a stand-alone answer about clinical relevance. Estimates and confidence intervals are needed to assess both direction and precision.

Why does crossover structure matter?

In general, crossover trials can use within-participant comparisons, which can account for stable differences between participants. Such analyses also need to address the structure of treatment periods and any relevant period or carryover effects. The registry identifies a 4-period crossover and lists treatment sequences, but the detailed statistical handling of period, sequence, or carryover effects is not reported in the registry record. The sponsor’s reported analyses are described as ANOVA, with least-squares mean differences. The precise model terms and assumptions are not reported in the registry record, so the model cannot be fully evaluated from the posted summary alone.

The group counts are 23 for placebo, 23 for donepezil, 25 for roflumilast, and 27 for the combination. The endpoint population is described as participants from the Full Analysis Set with data available for analysis at the given time-point. Since this is a crossover study, treatment-specific counts should not be added together as though they were necessarily separate people. The registry enrollment is 27, but the available outcome count can differ by treatment condition and analysis availability.

What do the standard deviations add?

A standard deviation describes variation in observed scores around a group mean. The posted standard deviations are 3.02 for placebo, 3.33 for donepezil, 3.62 for roflumilast, and 2.68 for the combination. Greater variation means participants’ changes were more spread out around the average; it does not indicate whether the mean change is better or worse. Standard deviations are not confidence intervals for the mean, and they should not be read as ranges containing all participants.

General statistical context: when outcomes are more variable, a study usually needs more information to estimate a mean difference with a given degree of precision. In a crossover study, the correlation between measurements from the same participant can also affect precision. That correlation is not included in the posted group summaries used for the simple recalculations. This distinction is one reason to give the sponsor’s reported model-based analysis priority when describing the trial result.

What is not specified?

The registry record does not report the original sample-size justification, a complete statistical analysis plan, the detailed ANOVA model, missing-data procedures, multiplicity adjustment, or interim analyses. These omissions limit assessment of whether the reported analysis followed a prespecified plan and how uncertainty was handled across the endpoint family. The registry states that the primary analysis population consists of Full Analysis Set participants with data available for the specified time-point, but it does not give the details needed to characterize imputation or sensitivity analyses. Such details should not be inferred from the outcome estimates alone.

10. Confidence Intervals and Precision

Precision is best assessed by considering the confidence interval alongside the point estimate. The sponsor’s reported 95% confidence interval for donepezil versus placebo is -1.4 to 0.8 correct responses, for roflumilast versus placebo it is -1.8 to 0.4, and for the combination versus placebo it is -1.0 to 1.2. Each interval includes zero, so the posted analysis does not clearly distinguish these estimated differences from no average difference.

The intervals also include values on both sides of zero. Consequently, the results do not establish the direction of a difference with confidence, nor do they rule out differences of some size in either direction. The endpoint’s possible score range is 0 to 18, but the registry does not report a threshold for a clinically important difference. It is therefore not possible to use the interval alone to determine whether the range of differences is clinically negligible or important.

The recalculated unadjusted intervals are wider than the sponsor’s reported intervals: -2.137 to 1.537 for donepezil versus placebo, -2.581 to 1.181 for roflumilast versus placebo, and -1.495 to 1.695 for the combination versus placebo. They use the posted group summaries and do not reproduce the sponsor’s ANOVA. The crossover structure matters because the relationship between repeated measurements on the same participant can inform precision, but that information is not present in the posted summary statistics. The recalculations are therefore not the preferred estimates for the trial’s reported inference.

Confidence intervals quantify uncertainty under the analysis model; they do not account for every possible source of bias or uncertainty in study conduct. The registry does not provide enough information to assess all model assumptions, missing-data handling, or multiplicity procedures. The intervals should be interpreted as uncertainty for the stated comparisons, not as proof of equivalence or as a measure of clinical importance.

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 illustrations based on the observed effect, not the trial’s original design justification. For these continuous-outcome comparisons, the figures are participants per arm, rather than total events. The stated planning assumptions are two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set at 100%, 75%, or 50% of the observed effect. For time-to-event outcomes, a planning display would instead give total events, but this primary endpoint is continuous.

The 80% power column describes the number of participants per arm that the planning calculation would require for an 80% chance of detecting the stated effect under the assumptions. The 90% power column uses a higher target chance of detection and therefore requires more participants per arm for the same assumed effect. For example, the planning figures for donepezil versus placebo are 1,763 participants per arm for 80% power and 2,360 for 90% power when planning for 100% of the observed effect. For the combination versus placebo, the corresponding figures are 12,670 and 16,961 participants per arm.

Smaller assumed effects need more participants because they are harder to distinguish from random variation. This is reflected in the planning scenarios using 75% and 50% of the observed effect, where the required numbers rise. The figures are highly dependent on the estimated effect used as the starting point, which is uncertain in a small trial. These calculations are not observed power; they are planning illustrations based on an effect estimate and stated assumptions. The original trial’s sample-size assumptions and justification are not reported in the registry record.

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 = -0.300; pooled standard deviation σ = 3.179.
2. n = 2 × (1.96 + 0.842)² × 3.179² / -0.300² = 1,762.4, rounded up to 1,763 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 <- -0.3                      # observed difference in means
sigma <- 3.17878121298085                     # 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(2051335)
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 = 1,763 (80% power), 2,360 (90% power)
 75% of observed effect: n per arm = 3,134 (80% power), 4,195 (90% power)
 50% of observed effect: n per arm = 7,050 (80% power), 9,438 (90% power)
power.t.test: 1763.4 
Simulated power at n = 1,763 per arm: 0.798 (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. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as ANOVA with least-squares mean differences and 95% confidence intervals. For the primary comparisons against placebo, the sponsor estimates are -0.31 for donepezil, -0.70 for roflumilast, and 0.12 for the combination. Their reported p-values are 0.573, 0.210, and 0.821. The confidence intervals for all three comparisons include zero.

The recalculations use the posted group means and standard deviations to form simple unadjusted comparisons. They estimate differences of -0.3, -0.7, and 0.1, respectively, with wider confidence intervals and different p-values. The small differences between point estimates do not make the methods interchangeable. In particular, the registry’s crossover design and the sponsor’s ANOVA mean that a reconstruction from group-level summaries cannot recover all information used in the reported analysis.

The discrepancy indicator is false, but the sponsor’s own analysis remains the appropriate primary reference for what the trial reported. The recalculated results are useful as a transparent comparison based on the posted summaries, not as a competing analysis that replaces the sponsor’s model. The registry record does not provide enough detail to reproduce the complete ANOVA specification or verify all of its assumptions.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary delayed verbal recall outcome and multiple secondary outcomes. These include immediate and delayed verbal recall at other time points, paired associates learning errors, rapid visual information processing signal detection and latency, spatial working memory errors, and safety outcomes. The registry reports 11 outcome measures and 10 analyses. Because multiple outcomes and comparisons are reported, multiplicity is relevant: when many statistical tests are conducted, the chance of at least one apparently unusual result can increase. Whether and how this was addressed is not reported in the registry record.

The sponsor’s reported analyses include comparisons of placebo with donepezil, roflumilast, and the combination, as well as comparisons between active conditions. The posted p-values are presented as reported, but an adjustment procedure is not specified. The unadjusted recalculated p-values are not adjusted for testing multiple comparisons. No conclusion about family-wise error control or a prespecified hierarchy of hypotheses can be drawn from the registry record.

The primary endpoint population is described as participants from the Full Analysis Set with data available for analysis at the given time-point. That definition indicates that the analysis uses available endpoint data, but the registry does not report how missing values were handled, why any data were unavailable, or whether sensitivity analyses were performed. Imputation methods, rules for excluding observations, and the extent of missingness are not reported in the registry record.

The original analysis plan, details of the Full Analysis Set definition beyond the posted description, and any interim analyses are not reported in the registry record. The same applies to detailed model terms for the crossover structure. Without these details, it is not possible to determine whether the reported analysis addresses period effects, sequence effects, carryover, or other features of the design, or whether the choices were prespecified. These gaps limit a full assessment of the analysis but do not change the values that the registry reports.

15. 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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: viral infection (4.3%).

The safety results give treatment-condition counts for serious and other adverse events. For serious adverse events, the posted counts are 1 in A: Placebo, 0 in B: Donepezil 10 mg, 0 in C: Roflumilast Dose A, and 0 in D: Roflumilast Dose A + Donepezil 10 mg. The corresponding numbers at risk are 23, 23, 25, and 27. These are the registry’s condition-specific figures; because the study is a crossover, they should not be interpreted as counts from independent parallel groups.

The registry identifies Viral infection, at 4.3%, among the top serious adverse event terms. The record does not assign that term to a treatment condition in the information. Counts for other adverse events are 7 for placebo, 9 for donepezil, 5 for roflumilast, and 12 for the combination, with condition-specific numbers at risk of 23, 23, 25, and 27. These are counts, not a basis for inferring comparative risk without additional information about event definitions, exposure, and analysis.

Death counts are not reported in the registry record. The registry also lists safety outcomes involving treatment-emergent adverse events, markedly abnormal laboratory tests, and markedly abnormal vital signs, but detailed results for those outcomes are not reported in the registry record for this analysis. The adverse-event figures describe what was posted and do not establish a complete safety profile.

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

The primary endpoint describes delayed verbal recall after scopolamine administration in healthy adults. Mean change was negative under placebo and each active condition, indicating worsening from baseline on this score according to the registry’s outcome definition. The posted group means were -6.7 for placebo, -7.0 for donepezil, -7.4 for roflumilast, and -6.6 for the combination. The sponsor’s primary comparisons against placebo have confidence intervals that include zero, so the registry results do not demonstrate a clearly separated mean change for any active condition on this endpoint.

That statistical reading is narrower than a claim that the treatments have no effect. A confidence interval spanning zero leaves uncertainty about the direction and size of a difference. The registry record does not report a clinically meaningful difference threshold, and the results do not establish equivalence. The findings also relate to a particular experimental setting: healthy adults, single oral administrations described in the official title, and scopolamine-induced deficits. They should not be generalized directly to treatment of Alzheimer’s disease, memory impairment in patients, repeated use, or outcomes not assessed by this endpoint.

The trial’s crossover design means that participants were studied across treatment periods, but the registry summary does not give enough detail to evaluate all design and analysis features. The posted means and sponsor’s model-based comparisons should be understood in that context. The unadjusted recalculations provide a separate view based on summary data; they do not replace the sponsor’s analysis.

The safety counts are descriptive and limited. The registry records one serious adverse event in the placebo condition and none in the other listed conditions, while also listing other adverse event counts. These figures alone do not establish comparative safety. Overall, the results describe uncertainty in the estimated cognitive outcome and a limited set of posted safety counts. They do not provide a broad clinical conclusion about efficacy or safety beyond the trial setting and outcomes recorded.

17. Why This Trial Matters Statistically

This trial is statistically interesting because it combines an experimentally induced cognitive deficit with a randomized crossover structure. The design can be useful for comparing treatment conditions within participants, since each participant may contribute observations under different conditions. That feature also makes the analysis dependent on the treatment-period structure and the relationship between repeated measurements. The registry identifies the crossover design and treatment sequences, but does not report enough model detail to assess how those features were handled.

The example also illustrates why a point estimate alone is not a complete result. The sponsor’s estimates are close to zero for the comparisons with placebo, but the confidence intervals include values on both sides of zero. The intervals communicate uncertainty about both direction and magnitude. A non-significant result is not proof of equivalence, particularly when the registry does not provide an equivalence margin or the original sample-size justification.

Another statistical lesson is the distinction between a sponsor-reported model and a simple reconstruction from group summaries. The latter can be informative as a check on the rough scale of the comparisons, but it does not retain the within-participant information that can matter in a crossover trial. The sponsor’s reported ANOVA is therefore the principal source for the trial’s inferential results, while the unadjusted recalculations should be treated as supplementary.

Finally, the registry records several cognitive and safety outcomes, creating questions about multiplicity and missing data. The reported primary analysis is interpretable at the level of its estimate and interval, but fuller assessment depends on analysis-plan details that are not reported. The record demonstrates how design, model specification, endpoint definition, and reporting completeness all affect what can be concluded from clinical trial statistics.

18. 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.

The registry record has limited information about the statistical plan. It does not report the original sample-size justification, detailed ANOVA model terms, multiplicity adjustment, missing-data procedures, or interim analyses. The method for addressing the crossover structure, including period, sequence, and carryover effects, is also not reported in the registry record. These omissions make it difficult to reproduce or fully assess the sponsor’s analysis.

The primary outcome is available for analysis in 23 participants in the placebo condition, 23 in the donepezil condition, 25 in the roflumilast condition, and 27 in the combination condition. The registry describes the analysis population as Full Analysis Set participants with available data at the relevant time-point, but does not explain the differences in available counts. The condition counts should not be added as if they necessarily represent separate participants because the design is crossover.

The recalculated comparisons use posted summary statistics and are unadjusted. They cannot incorporate within-participant covariance or reproduce the sponsor’s ANOVA specification. The sponsor’s own results should therefore lead interpretation. The reported primary confidence intervals include zero, while the registry does not provide a threshold for clinical importance or an equivalence margin.

The study concerns healthy adults with scopolamine-induced deficits and single administrations described in the official title. That context limits generalization to other populations, clinical conditions, treatment schedules, or outcomes. The posted outcome information for this analysis includes the primary endpoint and selected safety counts, but not detailed results for every secondary outcome. The registry’s posted safety figures are also descriptive and do not provide a complete comparative safety analysis.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Roflumilast and Donepezil to Reverse Scopolamine Induced Cognitive Deficits in Healthy Adults — NCT02051335.
• Registry record last updated 2017-02-01; status last verified 2016-09.
• Results first posted 2015-08-17.
• 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.

22. Record Summary

NCT02051335 is a completed Phase 1 randomized, double-blind, placebo-controlled, quadruple-masked, 4-period crossover study in healthy adults. The registry records enrollment of 27 and dates of 2014-01 for study start and 2014-05 for primary completion. Treatment conditions were placebo, donepezil 10 mg, roflumilast Dose A, and the combination, with scopolamine used to induce cognitive deficits.

The primary endpoint was change from baseline in delayed verbal recall at 1 hour after scopolamine. Mean changes were negative in all conditions. The sponsor’s ANOVA comparisons against placebo had 95% confidence intervals that included zero. The registry does not report enough detail to fully assess the analysis plan, missing-data handling, or multiplicity procedures. Safety counts include one serious adverse event in the placebo condition and none in the other listed conditions; deaths are not reported in the registry record.