Study to Demonstrate Cognitive Enhancing Effects of BF2.649: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study to Demonstrate Cognitive Enhancing Effects of BF2.649 (NCT00690274), a randomized phase 2 study. The primary outcome is Changes in Delayed Recall From Baseline to 12 Weeks.

2. Clinical Question

The registry describes the study as evaluating cognitive effects and symptom severity. The primary outcome focused on delayed recall, a particular aspect of memory assessed with the Brief Visuospatial Memory Test-Revised. Its reported t-scores are age-group based, and higher scores indicate recall of more items after a 45-minute delay. The secondary outcome concerned symptom severity between baseline and Week 12.

The outcome summaries describe the participants represented in each arm, but the registry record does not report a detailed analysis population or explain how participants with missing outcome measurements were handled. The number enrolled was 52; the primary outcome summaries list 23 participants in the placebo arm and 15 in the BF2.649 arm. The registry record does not explain the difference between enrollment and those outcome counts.

3. Trial Design

This was a randomized, double-blind, placebo-controlled, parallel-group phase 2 study. The registry lists the sponsor as University of Texas Southwestern Medical Center and records the study as completed. The listed enrollment was 52. The study start is recorded as 2008-06 and primary completion as 2011-10.

Participants were assigned to BF2.649 or placebo. The arm descriptions specify up to 20mg per day for each. The parallel design means the posted summaries compare separate groups rather than reporting a crossover in which every participant received both interventions. Randomization is intended, in general, to reduce systematic differences between groups at assignment. Double masking is intended to limit the influence of treatment knowledge on participants’ experiences and outcome assessment. The registry does not report who was masked or how masking was maintained.

The primary endpoint was change in delayed recall from baseline to 12 weeks. A secondary outcome addressed symptom severity over the same period. The registry record contains results for those outcomes but does not report a detailed protocol, eligibility criteria, allocation procedure, treatment duration beyond the outcome timeframe, adherence information, or the trial’s original sample-size assumptions. Those details should not be inferred from the posted outcome summaries.

The primary outcome lists 23 participants in the placebo group and 15 in the BF2.649 group. The secondary outcome lists the same arm sizes. These are outcome-specific counts in the posted summaries; the record does not state whether they represent all randomized participants, all participants with available measurements, or another analysis population.

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

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 Changes in Delayed Recall From Baseline to 12 Weeks, measured over 12 weeks.

The primary endpoint is titled “Changes in Delayed Recall From Baseline to 12 Weeks.” It uses the Delayed Recall component of the Brief Visuospatial Memory Test-Revised, a neuropsychological test of the ability to recall previously presented stimuli. The reported measure is a t-score, summarized as a group mean with a standard deviation.

According to the registry description, the reported t-score range is 35-81. Scores are calculated from individual scores with reference to each person’s age group. Higher scores indicate that more items were recalled after a 45-minute delay. Delayed recall is used to assess a specific memory function, rather than cognition in every domain. The endpoint therefore describes performance on this particular task and should not be treated as a direct measure of every aspect of cognitive functioning.

Comparing change from baseline to a specified follow-up makes the timing of the assessment explicit. In general, a mean summarizes the group’s central value, while the standard deviation describes the spread of individual values around that mean. The registry record does not provide further detail on the calculation of change scores, the treatment of missing measurements, or whether any adjustment was made for baseline scores. The posted result should therefore be interpreted in terms of the reported endpoint and group summaries, without assuming an analysis procedure that is not documented.

7. Results Posted in the Registry

• BF2.649: mean 43.40 versus 40.20 on control; difference 3.20 (95% CI 1.07 to 5.33), standardized effect d = 1.02; z-test p = 0.003.

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.

To Evaluate the Effect of BF2.649 on Symptom Severity Between Baseline and Week 12

• 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:
  • BF2.649: mean -4.90 versus 0.50 on control; difference -5.40 (95% CI -6.54 to -4.26), standardized effect d = -3.24; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for BF2.649 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry reports one secondary outcome: symptom severity between baseline and Week 12. Its posted group summaries and unadjusted comparison show a numerical contrast between BF2.649 and placebo, but the record does not identify the scale or explain how higher and lower scores correspond to symptom severity. The numerical direction therefore cannot be translated confidently into a direction of clinical change.

The secondary result is related to the trial’s stated aim of evaluating both cognitive effects and symptom severity. It addresses a different outcome domain from delayed recall, so it does not directly replicate or validate the primary memory result. Considered together, the outcomes describe two distinct questions asked over the same follow-up timeframe. The registry record does not provide enough information to determine whether the secondary measure was prespecified in a detailed protocol or how it was analyzed.

In general, secondary outcomes can provide useful context, but their statistical results often carry less weight than a prespecified primary outcome. A trial may examine multiple endpoints, time points, or analyses; the more comparisons made, the greater the chance that at least one result appears unusual by chance. An adjustment strategy or a hierarchy of outcomes can address this concern, but neither is reported in the registry record.

The secondary comparison is reported without adjustment, and the record lists no reported analyses. Its p-value should therefore be read as an unadjusted result, not as evidence corrected for the number of outcomes examined. The scale’s identity, score direction, analysis population, and missing-data handling are also not reported in the registry record. Those uncertainties limit the strength of conclusions that can be drawn from the secondary outcome and counsel against treating it as stand-alone confirmation of the primary finding.

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

• To Evaluate the Effect of BF2.649 on Symptom Severity Between Baseline and Week 12

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: each participant’s delayed-recall score is represented on a numeric scale, and the posted summaries report a mean and standard deviation for each arm. The comparison is expressed as a mean difference between BF2.649 and placebo. A difference of zero would mean equal group means on the reported scale; a positive difference means the BF2.649 mean is higher, and a negative difference would mean it is lower. For delayed recall, the registry says that higher t-scores indicate recall of more items after a 45-minute delay.

What does a mean summarize?

In general, a mean is the arithmetic average of the observed values in a group. It gives a compact description of the group’s central level, but it does not show every participant’s result. The standard deviation describes how dispersed the individual observations are around that mean. A small standard deviation indicates values closer to the mean than a larger standard deviation does, but it does not show the shape of the distribution or identify outliers. The primary outcome reports means and standard deviations, not individual participant measurements.

The reported group means are 40.2 for placebo and 43.4 for BF2.649. The standard deviations are 2.9 and 3.5, respectively. The registry identifies the endpoint as change from baseline to 12 weeks, but the record does not give the formula used to derive each change score or state whether the baseline value was included as a covariate in an adjusted model.

How should the mean difference be read?

The reported unadjusted mean difference is 3.2 t-score points for BF2.649 versus placebo. This is a contrast between group averages, not a change experienced by every participant. It is unadjusted in the sense that the posted comparison is presented without an adjustment for other variables or for multiple comparisons. The registry record does not report whether the investigators conducted additional adjusted analyses.

In general, an unadjusted comparison is a direct comparison of the groups’ reported summaries. It can be useful for understanding the observed contrast, but it may not account for baseline imbalance or other factors. Randomization is designed to support a fair comparison, yet it does not ensure identical groups in every realized trial, especially when group sizes are limited. The registry record does not provide baseline outcome values or other baseline characteristics with which to examine that issue.

What does the confidence interval add?

A confidence interval gives a range of values that reflects statistical uncertainty around an estimate under the assumptions of the calculation. The primary mean-difference interval is 1.069 to 5.331. It conveys more information than a single estimate because it indicates the range of differences that remain compatible with the data and method at the stated confidence level. A confidence interval is not a range containing a specified proportion of individual participant scores, nor does it guarantee that the true value lies within the particular interval.

General interpretation depends on how the interval was calculated, including the model and assumptions used. The record gives the interval but does not identify the test, variance method, or model behind the sponsor’s analysis. The posted comparison is described as a simple unadjusted recalculation from the summary statistics. Because the record reports no trial analyses, the calculation should not be mistaken for a fully documented investigator analysis.

What does the p-value say?

In general, a p-value measures how incompatible the observed data, or more extreme data, would be with a specified null model if that model and its assumptions were true. The primary unadjusted p-value is 0.003. It is not the probability that there is no difference, and it does not measure the probability that the result is due to chance in the ordinary sense. Nor does it indicate whether an effect is clinically important. The effect estimate and its interval are needed to understand magnitude and precision.

A p-value is sensitive to the number of observations and the variability in the data, as well as to the size of the estimated contrast. It cannot remedy limitations in trial conduct, outcome measurement, or analysis reporting. The registry does not report a multiplicity adjustment, so the posted p-value should be read as unadjusted rather than as a result corrected for examining multiple outcomes or analyses.

Why does over-dispersion matter in some analyses?

General statistical context: over-dispersion occurs when observed variability is greater than a model assumes. If a method underestimates variability, standard errors can be too small and confidence intervals too narrow; significance tests can then appear more compelling than warranted. This is especially relevant in some count-data models. The primary endpoint is reported as a continuous mean and standard deviation, and the registry record does not identify an over-dispersion model or report a diagnostic assessment. It would be inappropriate to claim that over-dispersion was present or that a particular correction was used.

The interpretation of any statistical result also depends on the analysis population and missing-data handling. The record reports arm-specific outcome counts but not how participants entered the analysis, whether every randomized participant had a measurement, or whether missing values were imputed. These are not minor details: different choices can affect both the estimate and its uncertainty. Without those details, the posted summaries can describe the included observations, but cannot fully document the trial’s inferential procedure.

11. Confidence Intervals and Precision

The primary estimate is a mean difference of 3.2 t-score points, with a 95% confidence interval from 1.069 to 5.331. The interval is entirely in the positive direction. In general, a confidence interval combines the estimate with its statistical uncertainty: a narrower interval indicates greater precision under the method’s assumptions, while a wider interval indicates less precision. It is not an interval for individual participants’ scores and does not capture every source of uncertainty, such as bias from missing data or limitations in the trial design.

The reported interval gives a range of plausible group differences under the statistical calculation used. Its lower and upper limits are not equally likely values in a probability distribution for the true effect. A 95% confidence procedure is interpreted over repeated applications of the same method, not as a guarantee that a particular interval contains the true value with a 95% probability.

The record does not identify the method used to calculate the sponsor’s interval, and it does not report a sponsor analysis. The unadjusted comparison is a simple recalculation from the posted summary statistics. The participant counts represented in the primary outcome are 23 for placebo and 15 for BF2.649. These group sizes, together with the observed variability, are relevant to precision, but the interval itself is the direct summary of uncertainty available in the record.

Precision should also be distinguished from practical or clinical importance. A confidence interval describes statistical uncertainty on the scale of the outcome; it does not supply a threshold for meaningful change. The registry record does not report a minimally important difference for delayed recall. It also does not show whether the result would be similar under a different population, outcome measure, or analysis strategy.

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 planning figures are participants per arm, not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect. The original trial’s own sample-size assumptions are not reported in the registry record.

The 80% power column gives an illustration of the number of participants per arm that would be needed, under those assumptions, for an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column gives the corresponding illustration for a 90% chance. These are design-planning quantities; they are not probabilities that the trial’s result is correct, and they do not describe the probability of detecting an effect after the trial has been completed.

The rows scale the effect used for planning relative to the observed effect: 100%, 75%, and 50%. As the assumed effect becomes smaller, more participants are needed to distinguish it from random variation under the same planning assumptions. The resulting figures should not be treated as the trial’s planned enrollment, because the registry does not report the original design assumptions. They are not observed power.

These illustrations depend on assumptions about allocation, significance level, and the approximation used. Different assumptions or a different target effect could produce different planning numbers. The record does not state whether the trial investigators used the observed effect, or any of these assumptions, when designing the study.

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 = 3.200; pooled standard deviation σ = 3.147.
2. n = 2 × (1.96 + 0.842)² × 3.147² / 3.200² = 15.2, rounded up to 16 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 <- 3.2                      # observed difference in means
sigma <- 3.14695620136877                     # 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(690274)
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 = 16 (80% power), 21 (90% power)
 75% of observed effect: n per arm = 27 (80% power), 37 (90% power)
 50% of observed effect: n per arm = 61 (80% power), 82 (90% power)
power.t.test: 16.2 
Simulated power at n = 16 per arm: 0.800 (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. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when investigators test more than one outcome, comparison, time point, or analysis. In general, conducting many tests increases the chance of at least one apparently notable result even when the underlying effects are absent. Statistical plans may address this through adjustments, a prespecified hierarchy, or a clear distinction between confirmatory and exploratory analyses. The registry record does not report a multiplicity strategy. The posted p-values are unadjusted, so they should not be interpreted as accounting for multiple testing.

The registry identifies a primary outcome and one secondary outcome, but it does not provide a full statistical analysis plan. The record also reports no analyses. As a result, it is not possible to determine from the record whether other outcomes, subgroups, time points, or models were considered, or whether the posted outcome summaries were selected from a wider set of analyses. This is a limit on interpretation, not evidence that additional analyses occurred.

Missing-data handling is also not reported in the registry record. The record gives 52 as enrollment and lists outcome summaries for 23 participants in the placebo group and 15 in the BF2.649 group. It does not explain whether the difference reflects unavailable measurements, exclusions, the analysis population, or another reason. No imputation method, sensitivity analysis, or accounting of missing observations is reported. It is therefore not possible to assess whether missingness could have influenced the estimates.

The record does not identify whether analyses followed an intention-to-treat principle, a per-protocol approach, or another population definition. Nor does it specify how randomization, baseline measurements, or protocol deviations were incorporated into the analysis. These methodological details affect what a group comparison estimates and how robust it may be. Without them, the posted statistics can be described, but the full inferential framework cannot be reconstructed.

These reporting gaps do not negate the numerical results. They define what can and cannot be concluded from the registry record. The estimates are best treated as summaries of the reported outcome data, with unadjusted comparisons and unresolved uncertainty about multiplicity, missingness, and analysis populations.

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

Serious events with the highest arm-level rate: suicidality (6.7%).

The registry reports serious adverse events affecting 0 participants in the placebo group and 1 participant in the BF2.649 group. It reports 0 participants with other adverse events in each group. These are counts among the listed at-risk groups of 23 for placebo and 15 for BF2.649. The record does not provide a death count; deaths are not reported in the registry record.

The registry lists suicidality at 6.7% among the top serious adverse event terms. The record does not make clear how that term relates to the group-level affected-participant counts, so it should not be assigned to a particular arm on this basis. The percentage is a reported term value, not a basis for inferring a comparative risk.

In general, adverse event counts describe events recorded during a study; they do not by themselves establish that an intervention caused an event. With small group counts, a single participant can substantially affect the observed pattern. The record does not provide event timing, severity details beyond the serious classification, exposure-adjusted rates, or a causal assessment. The figures therefore describe the safety reporting available in the registry, not a complete safety profile.

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

The primary result is a higher reported mean delayed-recall change in the BF2.649 group than in the placebo group. The estimated unadjusted difference is 3.2 t-score points, and the 95% confidence interval ranges from 1.069 to 5.331. Because the registry says that higher delayed-recall t-scores indicate recall of more items after a 45-minute delay, the numerical direction favors a higher score in the BF2.649 group on this particular measure.

That interpretation is limited to the measured outcome and the participants represented in the posted summaries. A delayed-recall score is not a complete measure of cognition, functioning, or everyday experience. The record does not report a threshold for meaningful change, so statistical evidence of a group difference cannot by itself establish whether the magnitude is important to participants or clinicians.

The result does not show that every participant in one group scored higher than every participant in the other. Group means summarize distributions that may overlap, and the standard deviations indicate that individual scores vary within each arm. The record does not provide participant-level data or a description of the score distributions.

The outcome summaries include 23 placebo participants and 15 BF2.649 participants, while enrollment is listed as 52. The registry does not explain the relationship between those counts or report how missing observations were handled. Nor does it report the trial’s original sample-size justification, a full analysis plan, a multiplicity strategy, or the analysis population. Those omissions limit conclusions about robustness and generalizability.

The secondary result addresses symptom severity, but the scale and its score direction are not reported. Its numerical contrast cannot therefore be given a confident clinical direction. More broadly, the registry results do not establish effects beyond the specific measures and timeframe recorded. They should be understood as statistical summaries from this completed phase 2 study, not as individual predictions or medical advice.

17. Why This Trial Matters Statistically

This trial illustrates why an effect estimate, its confidence interval, and the outcome definition should be read together. The estimated difference in delayed recall is expressed in the endpoint’s t-score units, and the interval communicates uncertainty around that group contrast. The score’s direction is defined in the registry: higher values indicate recall of more items after a 45-minute delay. This context gives the numerical estimate meaning without turning it into a broader claim about cognition.

The record also shows why outcome-level counts and overall enrollment should not be treated as interchangeable. Enrollment is 52, while the primary outcome summaries list 23 participants in the placebo arm and 15 in the BF2.649 arm. Without an explanation of the analysis population or missing-data handling, readers cannot determine why those counts differ. Transparent accounting is important for judging which participants contribute to an estimate.

Another statistical lesson is that a small unadjusted p-value does not answer every question. It does not quantify clinical importance, correct for multiple comparisons, or replace a confidence interval. The registry reports no multiplicity plan or trial analyses, so the posted result is most appropriately read as an unadjusted comparison rather than as a fully documented confirmatory analysis.

Finally, planning figures based on fractions of an observed effect can illustrate how sample-size requirements depend on the effect a future study aims to detect. They do not reconstruct the study’s original design or describe power observed after completion. Distinguishing planning illustrations from trial-design details prevents a retrospective calculation from being mistaken for the investigators’ rationale.

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 does not report detailed eligibility criteria, so the precise characteristics of the enrolled population cannot be established from the listed condition alone. The official title and summary refer to schizophrenia and schizoaffective disorder, while the condition list names schizophrenia. This leaves the composition of the study population unclear.

Outcome summaries list 23 participants in the placebo arm and 15 in the BF2.649 arm, compared with enrollment of 52. The record does not explain the difference, define the analysis population, or report how missing data were handled. Baseline values and other baseline characteristics are not included in the outcome summaries, limiting assessment of group comparability on measured characteristics.

The registry record does not report the original sample-size justification, the statistical analysis plan, the method used for the sponsor’s analysis, or any reported analyses. It also does not report a multiplicity strategy. The posted comparisons are unadjusted, and readers cannot establish from the record whether there were additional prespecified or exploratory analyses.

The primary measure is a specific delayed-recall test. The record does not report a threshold for meaningful change or broader cognitive outcomes that would establish how the score contrast relates to other domains. For the secondary symptom-severity result, the scale name and score direction are not reported, making clinical interpretation of the numerical direction uncertain.

Safety reporting is limited to participant counts and a listed serious-event term. The record does not report event timing, detailed descriptions, or an attribution assessment, and deaths are not reported in the registry record. These limitations make the registry summaries informative but incomplete for evaluating the full conduct, interpretation, and safety context of the study.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Study to Demonstrate Cognitive Enhancing Effects of BF2.649 — NCT00690274.
• Registry record last updated 2019-04-23; status last verified 2019-03.
• Results first posted 2019-04-23.
• 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

NCT00690274 is a completed randomized, double-blind, placebo-controlled, parallel-group phase 2 study sponsored by University of Texas Southwestern Medical Center. It enrolled 52 participants and compared BF2.649, up to 20mg per day, with placebo, up to 20mg per day, in people described as having schizophrenia or schizoaffective disorder.

The primary outcome was change in delayed recall from baseline to 12 weeks, assessed with the Brief Visuospatial Memory Test-Revised Delayed Recall t-score. The posted unadjusted comparison reports a mean difference of 3.2, with a 95% confidence interval from 1.069 to 5.331 and p = 0.003. One secondary symptom-severity outcome is also posted. The registry record does not report the trial’s analysis plan, multiplicity approach, sample-size rationale, analysis population, or missing-data handling.