A Study of VI-0521 for the Treatment of Obstructive Sleep Apnea / Hypopnea Syndrome in Obese Adults: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of VI-0521 for the Treatment of Obstructive Sleep Apnea / Hypopnea Syndrome in Obese Adults (NCT00745251), a randomized phase 1, phase 2 study. The primary outcome is Change in the Apnea/Hypopnea Index (AHI) Between Baseline and Week 28/Early Term..

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, placebo-controlled, parallel-group study listed in Phase 1 and Phase 2. The registry record identifies the condition as sleep apnea and describes the target population as obese adults with obstructive sleep apnea/hypopnea syndrome. The study is marked completed. Its recorded start was 2008-08 and its primary completion was 2009-09.

The enrollment was 45. The posted primary outcome results include 23 participants in the placebo arm and 22 in the Top Dose arm. The active arm is identified in the trial description as VI-0521 and in the arm description as 15 mg phentermine and 92 mg topiramate. The parallel-group structure means that participants were assigned to one of the two groups rather than crossing over between them; the registry does not report further details about the allocation process.

The primary outcome was the change in AHI between baseline and Week 28 or early termination, analyzed in an intent-to-treat population using last observation carried forward. The outcome is reported as a least-squares mean with standard error. A secondary outcome was percent change in weight from baseline to Week 28. The registry lists two outcome measures and posts one secondary outcome.

The registry record does not report the original sample-size justification, the planned statistical analysis in sufficient detail to reconstruct it, interim analyses, or details of random-sequence generation and masking procedures. Those omissions limit how fully the conduct and design rationale can be assessed from the record alone.

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:

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat Last-observation-carried-forward (ITT-LOCF)

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 the Apnea/Hypopnea Index (AHI) Between Baseline and Week 28/Early Term., measured over between baseline and Week 28.

The primary endpoint is change in the apnea/hypopnea index between baseline and Week 28 or early termination. AHI is the mean number of apnea or hypopnea episodes per hour of sleep, with episodes counted when they last at least 10 seconds. The unit, events/hour, expresses the frequency of these events relative to sleep time. A change from baseline summarizes how that frequency moved during the follow-up period for each participant.

AHI is used as a quantitative measure of sleep-disordered breathing because it combines apnea and hypopnea events into a rate that can be compared across participants and study groups. In general, analyzing change from baseline focuses on the movement in the measure during the study rather than only its final value. The posted group estimates are least-squares means, which are model-based means; the registry identifies the sponsor's comparison method as ANCOVA.

The reported primary analysis population is intent-to-treat with last observation carried forward. That label identifies the analysis population and missing-data approach recorded for the outcome, but the registry does not give enough detail to determine how many values were carried forward or what alternative missing-data analyses were considered. The endpoint therefore describes the posted AHI changes in the stated analysis population, not every possible way of analyzing follow-up data.

7. Results Posted in the Registry

• Top Dose: mean -31.46 versus -16.60 on control; difference -14.86 (95% CI -26.50 to -3.22), standardized effect d = -0.75; z-test p = 0.012.

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.

Percent Change in Weight From Baseline to Week 28

• 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 errors.
• Result:
  • Top Dose: mean -4.21 versus -10.26 on control; difference 6.05 (95% CI 2.83 to 9.27), standardized effect d = 1.10; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Top Dose than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcome is percent change in weight from baseline to Week 28. Across the two arms, the direction and relative ordering of the weight estimates do not match the ordering of the primary AHI changes. This contrast is relevant to the study's stated interest in the relationship between weight loss and OSA parameters, but an across-arm pattern alone cannot determine whether weight change explains the primary outcome difference.

The registry record does not report a formal mediation analysis, a statistical method for separating weight change from other pathways, or a prespecified hierarchy that would establish how secondary findings should be interpreted alongside the primary endpoint. The secondary endpoint therefore adds context but cannot, on its own, identify the mechanism behind the AHI contrast. Differences between estimates across outcomes are also not themselves evidence of a statistical interaction or causal relationship.

Unadjusted secondary results generally carry less evidential weight when multiple outcomes or comparisons are considered. General explanation: testing several outcomes creates more opportunities for apparently notable results to arise by chance unless the analysis plan addresses multiplicity. The registry does not report a multiplicity adjustment or a complete secondary-outcome analysis plan. The secondary result should consequently be treated as supportive or exploratory context rather than as an independent confirmation of the primary result.

The record lists two outcome measures and posts one secondary outcome. It does not report whether other planned outcomes were omitted, how the secondary analysis was selected, or whether its role was confirmatory. These gaps further limit how strongly the secondary result can be used to explain or reinforce the primary comparison.

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

• Percent Change in Weight From Baseline to Week 28

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: AHI is expressed as events/hour, and the analysis summarizes change from baseline. The registry reports least-squares means and standard errors for each arm. A least-squares mean is a model-based group mean, while a standard error describes uncertainty in estimating that mean. A standard error is not the same as the spread of individual participants' measurements. The record's sponsor analysis is identified as ANCOVA, a general method that compares outcome values between groups while accounting for other model terms, often including baseline outcome values. The specific covariates and model details are not reported in the registry record.

The sponsor reports a mean difference from ANCOVA for the primary outcome. Its estimate is -14.86, with p = 0.008. The posted confidence interval is incomplete because the lower limit is not reported in the registry record; the reported upper limit is -4.84. The incomplete interval should not be silently completed from another analysis. The sponsor's model result and the summary-statistic re-calculation answer related but not identical statistical questions because one is an ANCOVA result and the other is an unadjusted comparison.

The re-calculated primary contrast uses the posted arm means and sample sizes to form a simple unadjusted difference. The estimate is -14.86, and its reported recalculated 95% confidence interval is -26.5 to -3.218. Its unadjusted p-value is 0.012. The negative sign follows the contrast of Top Dose minus placebo: the Top Dose arm's estimated change is more negative. A confidence interval describes uncertainty around the estimated contrast under the analysis assumptions. A p-value describes how unusual data at least as incompatible with the null comparison would be under that null and the model assumptions; it is not the probability that the null hypothesis is true.

The re-calculated standard deviation is approximate, as indicated in the posted calculation. General explanation: where only arm-level standard errors and sample sizes are available, an approximate pooled standard deviation can be recovered under assumptions about the relationship between the standard errors and the group variability. That approximation is useful for an illustrative unadjusted calculation, but it does not reproduce the sponsor's participant-level ANCOVA. The registry record does not report participant-level data, model covariates, residual diagnostics, or enough information to verify the model assumptions directly.

The comparison also has a standardized effect estimate, which expresses the difference relative to a pooled measure of variability. In general, standardization can help describe magnitude in units of variability, but it does not replace the original clinical unit. For this endpoint, events/hour remains the directly interpretable scale. The pooled standard deviation reported for the re-calculation is 19.92, and the standardized estimate is negative because of the stated direction of subtraction. This re-calculated standardized quantity is not an additional sponsor analysis.

The displayed planning illustrations are derived from the observed effect, not from the trial's original design. General explanation: a power calculation asks how often a specified statistical procedure would detect an effect of a chosen size under stated assumptions if that effect were real. The calculation is sensitive to the assumed effect size, variability, allocation, significance level, and analysis model. The assumptions recorded for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to a stated fraction of the observed effect. These planning values are not a reconstruction of the original protocol's sample-size calculation, which is not reported in the registry record.

For a continuous outcome, the observed group mean difference can be compared with its uncertainty using a confidence interval or test statistic. General explanation: when variability is large relative to the difference, or when the sample is limited, estimates tend to be less precise. An interval helps show which values remain compatible with the analysis, subject to its assumptions. It does not incorporate every uncertainty about trial conduct, missing data, model choice, or generalizability.

Last observation carried forward is named as the missing-data approach for the primary analysis population. In general, this approach uses an earlier recorded value in place of a missing later value. Its consequences depend on why observations are missing and how values would have changed after the last measurement. The registry record does not report the amount or pattern of missing data, the timing of last observations, or sensitivity analyses using other assumptions. Therefore, the robustness of the primary estimate to missing-data choices cannot be assessed from the record alone.

11. Confidence Intervals and Precision

The re-calculated primary mean difference is -14.86 events/hour, with a 95% confidence interval from -26.5 to -3.218. The interval lies below zero, so the range of differences compatible with this unadjusted calculation is negative. It is nevertheless broad enough to include values with different magnitudes. The point estimate should therefore be read together with the interval rather than as an exact measure of the group contrast.

General explanation: a confidence interval is a range produced by a method that, over repeated comparable studies under its assumptions, would cover the target parameter at the stated rate. It is not a range containing a stated proportion of individual participant outcomes, nor is it a guarantee that the true value lies in this one interval. Its width reflects sampling uncertainty under the calculation, but not necessarily uncertainty from missing-data choices, model specification, measurement, or trial conduct.

The sponsor's ANCOVA analysis reports the same point estimate, -14.86, and an upper confidence limit of -4.84. Its lower confidence limit is not reported in the registry record, so a complete sponsor interval cannot be described. The sponsor's p-value is 0.008, while the simple unadjusted re-calculation has p = 0.012. Those values come from different analyses and should not be combined or treated as a single result.

The arm-level standard errors are 4.15 for placebo and 4.25 for Top Dose. They describe uncertainty in each estimated arm mean, not participant-to-participant variability. The re-calculation uses an approximate pooled standard deviation of 19.92; this reflects an additional assumption and does not turn the unadjusted interval into the sponsor's ANCOVA interval. The registry record does not report enough detail to quantify the precision of alternative analyses or the effect of missing observations.

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 event counts. They illustrate how many participants per arm would be needed under the stated assumptions to target 80% or 90% power for a chosen effect size. In plain language, the 80% power column describes a design that would have an 80% chance of detecting the specified effect at the stated significance level if that effect and the calculation assumptions were correct. The 90% column describes a higher target chance under the same kind of assumptions.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each illustration sets the effect to detect as a stated fraction of the observed effect. The displayed scenarios therefore use the observed result as an input; they are not the original trial's design justification, which is not reported in the registry record.

As the effect to be detected becomes smaller, more participants per arm are needed to distinguish it from random variation under the same assumptions. The rows for 100%, 75%, and 50% of the observed effect illustrate that relationship. The figures are planning illustrations, not observed power: observed power is not a useful substitute for interpreting the estimate and its confidence interval. The calculation also does not establish that the assumptions match the trial's original design or analysis plan.

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 = -14.860; pooled standard deviation σ = 19.918.
2. n = 2 × (1.96 + 0.842)² × 19.918² / -14.860² = 28.2, rounded up to 29 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 <- -14.86                      # observed difference in means
sigma <- 19.9181231020138                     # 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(745251)
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 = 29 (80% power), 38 (90% power)
 75% of observed effect: n per arm = 51 (80% power), 68 (90% power)
 50% of observed effect: n per arm = 113 (80% power), 152 (90% power)
power.t.test: 29.2 
Simulated power at n = 29 per arm: 0.796 (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 sponsor reports an ANCOVA comparison for the primary outcome with a mean difference of -14.86 and p = 0.008. The posted sponsor confidence interval has an upper limit of -4.84; its lower limit is not reported in the registry record. The simple re-calculation from the arm summary statistics also estimates -14.86, but gives an unadjusted p-value of 0.012 and a complete recalculated interval from -26.5 to -3.218.

The matching point estimates do not make the analyses identical. ANCOVA is the sponsor's reported model-based analysis; the re-calculation uses posted group summaries without reproducing that model. Differences in the p-values and interval reporting are therefore not, by themselves, evidence of a contradiction. The registry marks the discrepancy status as false, and the primary point estimates agree. The sponsor analysis remains the trial's reported analysis; the re-calculation is an illustrative unadjusted check rather than a replacement for the participant-level analysis.

For the secondary weight outcome, the sponsor's ANCOVA estimate uses the opposite contrast direction from the re-calculated Top Dose-minus-placebo difference. The magnitude corresponds, but the sign differs because of the contrast orientation. That distinction matters when interpreting whether a positive or negative value indicates a larger change in one arm.

15. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the primary and secondary analyses were protected by a prespecified multiplicity strategy, such as an adjustment to account for testing more than one outcome. Without that information, the unadjusted secondary p-value should not be interpreted as though it necessarily had the same confirmatory status as a single prespecified primary test. The record lists two outcome measures and posts one secondary outcome, but does not describe the broader analysis hierarchy or whether additional tests were planned.

The primary analysis population is described as intent-to-treat with last observation carried forward. In general, intent-to-treat analysis aims to retain participants according to their randomized groups, while last observation carried forward is a particular way to fill in missing follow-up values. The registry does not report the amount of missingness, reasons for missing observations, whether all randomized participants contributed to the outcome analysis, or whether sensitivity analyses were conducted. It is therefore not possible to assess from the record how strongly the estimate depends on the missing-data approach.

General explanation: carrying the last observed value forward assumes, in effect, that this value is an adequate stand-in for the unobserved later outcome. That assumption can be questionable when outcomes change over time or when missingness is related to a participant's condition or response. Alternative missing-data methods can make different assumptions and yield different estimates. No alternative analysis is reported in the registry record.

The sponsor's primary comparison is described as ANCOVA, but its covariates, model diagnostics, and handling of missing values beyond the named analysis population are not reported. Interim analyses, stopping rules, prespecified analysis details, and the original sample-size rationale are also not reported in the registry record. These omissions do not establish that such methods were absent; they mean that their use and consequences cannot be evaluated from the posted record.

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 4% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cerebrovascular accident (4.3%).

The registry reports serious adverse events affecting 1 of 23 participants in the placebo group and 0 of 22 in the Top Dose group. The listed serious event term is cerebrovascular accident, reported as 4.3%. The registry information presented does not identify the arm associated with that term, so the term should not be assigned to either group from this record alone. Death counts are not reported in the registry record.

Other adverse events are reported for 17 of 23 participants in the placebo group and 20 of 22 in the Top Dose group. These are separate from the serious adverse-event counts and should not be combined with them. The counts describe participants with reported events; they do not provide event severity, timing, duration, recurrence, or a complete account of individual clinical circumstances.

General explanation: small group counts can be unstable, and the absence of a reported serious event in one group does not establish that an event cannot occur. These registry figures describe the posted safety information but do not, by themselves, establish comparative safety. More detailed safety interpretation is limited because the record does not report deaths as counts and does not provide a complete narrative or participant-level listing of events.

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

The primary result shows that the posted mean AHI change was more negative in the Top Dose arm than in the placebo arm, with an estimated unadjusted difference of -14.86 events/hour. The interval for that re-calculation is entirely below zero, while still allowing a range of possible magnitudes. The sponsor's ANCOVA analysis reports the same point estimate and a p-value of 0.008, although the lower confidence limit is not reported.

These results describe the statistical comparison in this trial; they do not establish that every participant experienced a similar change or that the estimated difference has a particular clinical importance. The registry record does not report a threshold for a clinically important AHI change. Statistical evidence of a group difference and clinical interpretation are related but distinct questions.

The secondary weight outcome provides context but does not establish that weight change caused, mediated, or fully explained the AHI contrast. The registry does not report a formal analysis of that relationship. Nor do the posted outcome results answer questions about long-term outcomes beyond the stated follow-up or about populations not represented in this study.

The adverse-event counts are limited in scope and detail. They should be read as the events recorded in the registry, not as a complete characterization of risk. Missing-data handling, limited information on model specification, and the absence of a reported multiplicity plan also constrain interpretation. Taken together, the posted results support a description of the observed trial estimates and their statistical uncertainty, rather than a broad conclusion about treatment effects outside this study.

18. Why This Trial Matters Statistically

This trial illustrates why an estimated difference, its uncertainty, and the analysis method should be considered together. The primary point estimate is the same in the sponsor's ANCOVA analysis and in the simple re-calculation, yet their p-values differ and the sponsor's confidence interval is incomplete in the registry record. The distinction makes clear that a point estimate alone does not reveal how the comparison was modeled or how precisely it was estimated.

The endpoint also highlights the value of preserving measurement units. AHI is reported in events/hour, which gives the difference a direct interpretation as a change in event frequency relative to sleep time. Standardized measures and p-values can add statistical context, but they do not replace the original outcome scale or answer whether a difference is clinically important. The registry does not specify a threshold for that judgment.

The study's stated interest in weight loss and OSA parameters raises a further statistical issue: observing changes in two related outcomes does not identify the pathway connecting them. Establishing whether weight change accounts for a portion of an AHI difference requires an appropriate analysis, not simply juxtaposing outcome estimates. Such an analysis is not reported in the registry record.

Finally, the record demonstrates how important transparency is for judging evidence. Information about multiplicity, missingness, model details, and the original design rationale would help readers assess robustness and interpret planning choices. Where these details are not reported, statistical conclusions should remain bounded by the posted estimates and the assumptions that can be identified.

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.

The sample comprised 45 participants, with 23 in the placebo arm and 22 in the Top Dose arm. Estimates from groups of this size can be imprecise, as reflected in the interval for the unadjusted primary contrast. The registry does not report a clinically important difference threshold, so the magnitude cannot be classified as clinically meaningful using a criterion in the record.

The primary analysis used intent-to-treat last observation carried forward, but the record does not give the amount or pattern of missing data or report sensitivity analyses. The consequences of this approach therefore cannot be assessed. The sponsor's ANCOVA is named, but its covariates and further model details are not reported. The sponsor confidence interval also lacks a reported lower limit.

The registry does not report a multiplicity adjustment or a complete analysis hierarchy. This limits interpretation of secondary findings, particularly as confirmatory evidence. Although the study objectives include assessing the relative contribution of weight loss to OSA parameters, no formal analysis separating that contribution is reported.

Further details about random-sequence generation, masking procedures, the original sample-size justification, and interim analyses are not reported in the registry record. Their absence from the record does not prove that these procedures were not used, but prevents assessment of them. Safety information is also limited: death counts are not reported, and the serious-event term is not linked to an arm in the posted detail. These limitations restrict conclusions to the trial's reported estimates and event counts.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of VI-0521 for the Treatment of Obstructive Sleep Apnea / Hypopnea Syndrome in Obese Adults — NCT00745251.
• Registry record last updated 2012-10-05; status last verified 2012-09.
• Results first posted 2012-10-05.
• 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

NCT00745251 is a completed randomized, double-blind, parallel-group Phase 1 and Phase 2 study of VI-0521 versus placebo in obese adults with obstructive sleep apnea/hypopnea syndrome. The registry reports enrollment of 45 and records the study start as 2008-08 and primary completion as 2009-09.

The primary outcome was change in AHI from baseline to Week 28 or early termination, analyzed in an intent-to-treat last-observation-carried-forward population. The sponsor reports an ANCOVA mean difference of -14.86 and p = 0.008. The unadjusted re-calculation gives the same estimate, with a 95% confidence interval from -26.5 to -3.218 and p = 0.012.

The posted secondary outcome is percent change in weight from baseline to Week 28. The record also reports serious and other adverse-event counts, while deaths are not reported. Details on multiplicity, missingness, model specification, and the original sample-size rationale are not reported in the registry record.