Efficacy and Safety of S-Equol on Men With Benign Prostatic Hyperplasia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of S-Equol on Men With Benign Prostatic Hyperplasia (NCT00962390), a randomized phase 2 study. The primary outcome is Change From Baseline at Week 4 in Prostate Specific Antigen (PSA) Concentration..

2. Clinical Question

3. Trial Design

This was a Phase 2, randomized, double-blind, multicenter, placebo-controlled, parallel-group proof-of-concept trial. It evaluated 4 weeks of treatment with S-equol in men with benign prostatic hyperplasia. The listed interventions were S-equol and placebo, and the four primary-endpoint groups were S-equol 10 mg BID, S-equol 50 mg BID, S-equol 150 mg BID and placebo BID. The registry record reports enrollment of 116 participants and a completed status.

The study began in 2009-06, and primary completion was recorded as 2016-01. The primary endpoint was change from baseline at Week 4 in PSA concentration. The primary-endpoint population was defined as participants who received at least one dose of study drug and had a post-dose PSA assessment at Week 4. The posted analysis counts were 25 for the 10 mg group, 26 for the 50 mg group, 24 for the 150 mg group and 29 for placebo. These are endpoint analysis counts, not a description of the full enrollment process.

The registry lists 13 secondary outcomes as posted, including measures of prostate volume, urinary flow, void volume, post-void residual volume, dihydrotestosterone and luteinizing hormone. The record lists 14 outcome measures overall and 3 analyses. Further details on the number or location of centers, allocation concealment, the randomization procedure, treatment adherence and the original statistical analysis plan are not reported in the registry record. The trial description and results therefore support interpretation of the posted comparisons, but not reconstruction of every operational or analytic decision.

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: All Participants who received at least 1 dose of study drug and who had a post-dose PSA assessment at Week 4.

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.

Population: Participants who took at least one (1) dose of investigational product were included in the baseline analysis population.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline at Week 4 in Prostate Specific Antigen (PSA) Concentration., measured over 4 weeks.

The primary endpoint was change from baseline at Week 4 in prostate specific antigen (PSA) concentration, measured in ng/mL. PSA is a continuous laboratory measure. The posted group summaries are least-squares means, with standard errors as the reported measure of dispersion. A negative change indicates a lower Week 4 value relative to baseline in the group summary; it does not, by itself, establish why the measure changed or whether a change is clinically meaningful.

The registry rationale describes PSA as the most sensitive measure of S-equol effects on the prostate because of expected effects on the androgen receptor axis. That rationale explains why the endpoint was selected for this proof-of-concept study. It should be read as the study’s stated basis for measuring PSA, not as evidence that PSA captures every relevant effect or that a particular PSA change corresponds to a patient-important outcome.

The primary analysis population included participants who received at least one dose and had a post-dose PSA assessment at Week 4. The posted least-squares mean changes were -0.1 ng/mL for S-equol 10 mg BID, -0.1 ng/mL for S-equol 50 mg BID, -0.3 ng/mL for S-equol 150 mg BID and -0.4 ng/mL for placebo BID. The comparison of interest is the difference in change between each dose group and placebo, with uncertainty around that difference considered alongside the point estimate.

7. Results Posted in the Registry

• S-equol 10 mg BID: mean -0.10 versus -0.40 on control; difference 0.30 (95% CI -0.53 to 1.13), standardized effect d = 0.19; z-test p = 0.480.
• S-equol 50 mg BID: mean -0.10 versus -0.40 on control; difference 0.30 (95% CI -0.53 to 1.13), standardized effect d = 0.19; z-test p = 0.480.
• S-equol 150 mg BID: mean -0.30 versus -0.40 on control; difference 0.10 (95% CI -0.73 to 0.93), standardized effect d = 0.06; z-test p = 0.814.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous, meaning that it is measured on a numerical scale rather than as a category. For a continuous outcome, a group mean summarizes the central tendency of the observed values. The registry reports least-squares mean changes for PSA, together with standard errors. A least-squares mean is a model-based mean; the registry identifies ANCOVA for the reported comparisons, but the specific model terms and covariates are not reported in the registry record.

What does ANCOVA compare?

Analysis of covariance, or ANCOVA, uses a statistical model to compare groups while accounting for covariates included in that model. In a baseline-adjusted analysis, baseline outcome values are often considered, but the exact covariates used in this trial are not reported in the registry record. The results are described as mean differences in final values, so they should not be conflated with the simple subtraction of the two posted change means. The reported ANCOVA estimates are the trial’s stated comparisons.

What is a mean difference?

A mean difference expresses the estimated separation between a treatment group and the comparator on the outcome scale. Here, the posted sponsor estimates are 0.3 ng/mL for S-equol 10 mg BID versus placebo, 0.3 ng/mL for S-equol 50 mg BID versus placebo and 0.1 ng/mL for S-equol 150 mg BID versus placebo. The positive estimates indicate that the treatment-group final values were higher relative to placebo under the reported comparison, but the confidence intervals are wide enough to include differences below and above zero. A point estimate should therefore be read together with its interval, not as a complete description of the result.

What does a standard error describe?

A standard error describes the estimated sampling uncertainty of a statistic, such as a mean. It is not the same as the spread of individual participant values. The arm summaries report a standard error of 0.3 for each group. The standard error helps describe precision of the group mean estimate; it does not show the full distribution of individual PSA changes or the range of values among participants.

How should p-values be read?

A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be under the assumptions of the statistical test. It does not give the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The approximate unadjusted p-values are 0.48 for the 10 mg comparison, 0.48 for the 50 mg comparison and 0.814 for the 150 mg comparison. These are not the sponsor’s reported ANCOVA p-values; those are reported to more decimal places than permitted for presentation. The registry’s reported confidence intervals and rounded effect estimates remain useful for interpreting the sponsor analysis.

Why use a confidence interval?

A confidence interval expresses uncertainty about an estimated effect under the assumptions of the analysis. A 95% confidence procedure is designed so that, over repeated applications under its assumptions, intervals would contain the target parameter at the stated frequency. It is not a statement that there is a 95% probability that this particular interval contains the true effect. For the posted sponsor comparisons, each 95% interval includes zero and includes values on both sides of zero. That indicates limited precision for distinguishing the direction and magnitude of the group difference.

What is different about the re-calculation?

The unadjusted re-calculations are derived from posted summary statistics, rather than the trial’s full ANCOVA model. The pooled standard deviations used for these calculations are marked as approximate, and the posted group dispersions are standard errors rather than standard deviations. Reconstructing variability from summary values requires assumptions and does not reproduce individual-level model fitting. The re-calculated estimates are best treated as an independent, approximate comparison, not as a correction to the registry analysis.

Why can over-dispersion matter?

Over-dispersion occurs when observed variability is greater than a statistical model expects. In general, ignoring over-dispersion can make standard errors too small and confidence intervals too narrow, creating a misleading impression of precision. The primary endpoint is continuous, and the registry record does not report a count model or an over-dispersion assessment. This general concept is relevant to statistical modeling broadly, but the record does not establish that over-dispersion occurred in this trial.

What does the sample-size illustration assume?

The planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. It uses a stated fraction of the observed effect as the effect to detect. It gives participants per arm for non-time-to-event outcomes such as this continuous endpoint. These are planning calculations based on the observed effect, not the trial’s original sample-size calculation. The original design assumptions beyond the rationale in the registry description are not reported in the registry record.

10. Confidence Intervals and Precision

The sponsor’s ANCOVA estimates are accompanied by 95% confidence intervals. For S-equol 10 mg BID versus placebo, the estimated mean difference in final values was 0.3, with an interval from -0.5 to 1.1. For S-equol 50 mg BID versus placebo, the estimate was 0.3, with an interval from -0.5 to 1.1. For S-equol 150 mg BID versus placebo, the estimate was 0.1, with an interval from -0.8 to 1.0. Each interval includes zero and spans values on either side of it.

The point estimates alone can make the comparisons appear straightforward, but the intervals convey that the data are compatible with a range of possible differences. The intervals do not support a precise conclusion about the direction or size of the difference for any dose comparison. They also do not establish equivalence: the registry does not report an equivalence margin or a non-inferiority margin, and the intervals are not interpreted against such a prespecified threshold.

Precision depends on the amount of information and the variability in the observations, as well as the analysis method. The reported standard errors for the arm means are 0.3. The endpoint counts range from 24 to 29 across the posted groups. Those counts and standard errors provide context for why the estimates should not be treated as exact. The record does not provide individual-level data or additional uncertainty measures that would permit a fuller assessment of the outcome distributions.

The approximate unadjusted comparisons also have broad uncertainty, but their reconstructed intervals include values to four decimal places and are not repeated. Their estimates and p-values are only a secondary check based on summary statistics. For interpretation of the trial, the sponsor’s reported ANCOVA estimates and rounded confidence intervals are the principal posted analysis results.

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 sample-size planning figures are expressed as participants per arm, not as total participants. This is a continuous endpoint, not a time-to-event outcome; therefore the display refers to participants per arm rather than total events. Its assumptions are a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. For each comparison, the effect to detect is the stated fraction of the observed effect. The trial’s own full design assumptions are not reported in the registry record. The registry’s endpoint description does state that 26 subjects in each treatment arm was considered adequate to observe a trend in this proof-of-concept study, and describes an estimate of 124 male subjects to achieve approximately 104 completed subjects based on an estimated drop-out rate of 15%.

The 80% power column gives the number of participants per arm that the planning calculation estimates would be needed for an 80% chance of detecting the selected effect under its assumptions, if that effect were the true one. The 90% power column uses a higher target chance of detection and consequently gives larger planning counts. These are planning illustrations, not observed power.

The calculation for the 10 mg comparison gives 427 participants per arm for 80% power and 571 for 90% power when planning for 100% of the observed effect. For 75% of the observed effect, the figures are 758 and 1,015; for 50%, they are 1,706 and 2,283. For the 50 mg comparison, the corresponding pairs are 434 and 580, 770 and 1,031, and 1,733 and 2,319. For the 150 mg comparison, they are 3,779 and 5,059, 6,718 and 8,993, and 15,115 and 20,234. Smaller effects require more participants to detect under the same assumptions.

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 = 1.563.
2. n = 2 × (1.96 + 0.842)² × 1.563² / 0.300² = 426.3, rounded up to 427 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 <- 1.56328057425599                     # 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(962390)
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 = 427 (80% power), 571 (90% power)
 75% of observed effect: n per arm = 758 (80% power), 1,015 (90% power)
 50% of observed effect: n per arm = 1,706 (80% power), 2,283 (90% power)
power.t.test: 427.2 
Simulated power at n = 427 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.

13. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparisons as ANCOVA estimates of mean difference in final values. These estimates are 0.3 for S-equol 10 mg BID versus placebo, 0.3 for S-equol 50 mg BID versus placebo and 0.1 for S-equol 150 mg BID versus placebo. The corresponding posted 95% confidence intervals are -0.5 to 1.1, -0.5 to 1.1 and -0.8 to 1.0.

The approximate re-calculations from posted summary statistics give differences of 0.3, 0.3 and 0.1, respectively. They are unadjusted and use approximations to reconstruct variability; they are not a re-fit of the ANCOVA model. Small differences between methods can arise because the calculations do not use identical model inputs. The record marks no discrepancy between the posted and recalculated results. The sponsor’s reported ANCOVA remains the trial’s own analysis, while the re-calculation serves as an approximate, independent check.

The sponsor p-values are present in the registry to more decimal places than can be reproduced under the presentation limit. The rounded effect estimates and confidence intervals are sufficient to see that all three intervals include zero. The registry record does not report further detail on ANCOVA covariates, model diagnostics or any adjustment to the p-values for multiple comparisons.

14. Multiplicity, Missing Data and Analysis Populations

The primary endpoint was PSA change at Week 4, and the registry lists three comparisons of S-equol doses with placebo. It also lists 13 posted secondary outcomes, including changes in prostate volume, Qmax, void volume, post-void residual volume, dihydrotestosterone concentration and luteinizing hormone concentration. Testing several doses and outcomes can increase the chance of at least one apparently unusual result arising by chance when each test is considered separately. This is the general statistical issue of multiplicity.

The registry record does not report whether the primary comparisons were adjusted for multiplicity, whether a testing hierarchy was prespecified, or whether any secondary outcomes were designated as confirmatory. The posted unadjusted re-calculations should not be interpreted as multiplicity-adjusted evidence. The sponsor analyses are identified as ANCOVA, but the record does not describe a family-wise error procedure or another adjustment for the dose comparisons. Accordingly, the p-values should be read as comparison-specific results, not as evidence that multiplicity has been controlled.

Missing data can also affect a treatment comparison if outcome availability differs across groups or relates to participants’ outcomes. The primary analysis population requires a dose of study drug and a post-dose PSA assessment at Week 4, which means the endpoint summaries pertain to that defined population. The registry record does not report how missing Week 4 PSA values were handled, whether imputation was used, the reasons for missing assessments, or sensitivity analyses for missing data. It also does not describe the relationship between the enrollment count and the endpoint analysis population in enough detail to reconstruct participant flow.

The registry’s primary-endpoint population definition is useful, but it is not equivalent to a complete missing-data analysis. It does not establish whether the analysis followed an intention-to-treat principle, nor does it describe a per-protocol population. The record does not report further analysis-population definitions or the detailed statistical plan. These omissions limit assessment of how robust the comparisons are to missing outcomes and analytic choices.

15. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: blood pressure increased (3.6%), electrocardiogram abnormal (3.6%), heart rate increased (3.6%), dyspnoea (3.4%).

The serious adverse event figures are reported for the safety groups, with the number at risk and the number affected listed separately. In S-equol 10 mg BID, 1 participant among 28 at risk had a serious adverse event; in S-equol 50 mg BID, none among 29 at risk were affected; in S-equol 150 mg BID, 1 among 29 at risk was affected; and in placebo BID, none among 29 at risk were affected. The registry reports zero deaths in each group.

The listed serious adverse event terms include Blood pressure increased, Electrocardiogram abnormal and Heart rate increased, each reported as 3.6%, and Dyspnoea, reported as 3.4%. These percentages are registry entries for the listed terms; the record does not provide enough detail in this summary to infer more about their timing, severity, relation to treatment or overlap among participants.

These counts are descriptive safety findings, not a statistical demonstration that groups have the same or different risk. With few reported serious events, the figures are imprecise. The absence of deaths in the posted groups does not establish absence of other harms. The record’s adverse-event information should be read as a limited account of events during the study, not as a complete safety profile for use beyond the trial conditions.

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

The posted result is a comparison of short-term PSA measurements in a randomized, double-blind, placebo-controlled proof-of-concept study. At Week 4, all four group summaries show negative PSA changes. The sponsor’s ANCOVA estimates for the S-equol dose groups versus placebo are 0.3, 0.3 and 0.1 ng/mL, with 95% confidence intervals that include zero. The data therefore do not precisely determine the direction or size of the group differences.

This statistical uncertainty is not proof that the treatments and placebo are identical. It is also not evidence that a difference exists. The confidence intervals include both negative and positive values, and the registry does not specify a threshold that would distinguish clinically important from unimportant changes. The result should be described as an imprecise estimate, rather than translated into a claim of benefit, no benefit or equivalence.

The endpoint is PSA change over 4 weeks, and the registry’s rationale for choosing it concerns expected effects on the prostate and androgen receptor axis. A result on this laboratory endpoint does not by itself establish an effect on symptoms, prostate volume, urinary function, long-term outcomes or quality of life. The registry lists several secondary outcomes, but their results are not part of the primary comparisons. The primary result therefore answers a narrower question than the overall clinical question of how treatment affects men with benign prostatic hyperplasia.

The study’s proof-of-concept description is also important. The observed arm counts for the primary endpoint are 25, 26, 24 and 29. The uncertainty around the estimates, together with the short treatment timeframe, constrains what can be inferred from the trial. Further information about the protocol, missing-data handling, model specification and multiplicity procedures is not reported in the registry record. These limits should accompany any interpretation of the posted results.

17. Why This Trial Matters Statistically

This trial offers a useful statistical example of why a point estimate should not be read alone. The reported differences between S-equol dose groups and placebo are small in absolute terms, while the 95% confidence intervals cover values on both sides of zero. That combination makes precision central to interpretation: the observed estimates are not enough to establish a clear direction or a narrowly bounded effect.

The study also illustrates the distinction between a trial’s reported analysis and a summary-statistic re-calculation. The registry identifies ANCOVA as the sponsor’s method. Approximate unadjusted calculations can check whether the posted summaries tell a broadly similar story, but they cannot recreate the analysis model or replace the sponsor’s reported estimates. The distinction is particularly important when model covariates and individual-level data are not available.

For planning, the observed differences produce large sample-size illustrations, especially when the target effect is set to a smaller fraction of the observed effect. This demonstrates a general statistical point: designing a study to detect a smaller effect generally requires more participants under the same assumptions. The figures are planning scenarios rather than a measure of the trial’s observed power, and the registry’s original design assumptions are not fully reported.

Finally, the record shows why results need to be read in context. The primary outcome is a short-term laboratory measure; multiple doses and numerous secondary outcomes are listed; and details on missing data and multiplicity are not reported. Statistical interpretation requires attention to the endpoint, uncertainty, analysis population and limits of the record, rather than relying on a single estimate or p-value.

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 primary endpoint reflects PSA change at Week 4, so the posted results do not describe longer-term changes. The registry lists secondary outcomes, but those results are not included in the primary comparisons. No threshold for a clinically important PSA difference is reported in the registry record, so the clinical relevance of a change cannot be determined from these estimates alone.

The endpoint analysis includes participants who received at least one dose and had a post-dose PSA assessment at Week 4. The registry does not report the missing-data method, reasons for missing assessments or sensitivity analyses. It also does not report the complete participant flow needed to understand all differences between enrollment and endpoint counts.

The sponsor’s comparisons are identified as ANCOVA, but the covariates, detailed model specification, diagnostics and handling of missing data are not reported in the registry record. The approximate re-calculations use posted summary statistics and approximate variability; they cannot reproduce individual-level adjusted analyses. No multiplicity procedure is reported despite multiple dose comparisons and numerous listed secondary outcomes.

The primary-endpoint group counts are 25 for S-equol 10 mg BID, 26 for S-equol 50 mg BID, 24 for S-equol 150 mg BID and 29 for placebo BID. The confidence intervals are broad relative to the point estimates and include zero for every comparison. The study’s proof-of-concept description and 4-week treatment duration further constrain the scope of inference. Details such as the full eligibility criteria, center locations, adherence, allocation concealment and original analysis plan are not reported in the registry record. These limitations do not negate the posted results, but they limit how far those results can be generalized or interpreted.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Efficacy and Safety of S-Equol on Men With Benign Prostatic Hyperplasia — NCT00962390.
• Registry record last updated 2017-05-19; status last verified 2017-04.
• Results first posted 2017-05-19.
• 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

NCT00962390 is a completed Phase 2 randomized, double-blind, multicenter, placebo-controlled, parallel-group proof-of-concept trial of S-equol in men with benign prostatic hyperplasia. Enrollment was 116, and the treatment period was 4 weeks. The primary endpoint was change from baseline at Week 4 in PSA concentration.

The posted ANCOVA estimates versus placebo were 0.3 for S-equol 10 mg BID, 0.3 for S-equol 50 mg BID and 0.1 for S-equol 150 mg BID. Each reported 95% confidence interval includes zero. Approximate unadjusted re-calculations give similar point estimates but do not replace the sponsor’s analysis. The posted serious adverse event counts were low, with zero deaths reported in each group. The registry record does not report important details about missing-data handling, multiplicity adjustment or the full model specification.