Safety and Efficacy of Bexagliflozin as Monotherapy in Patients With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Bexagliflozin as Monotherapy in Patients With Type 2 Diabetes (NCT02715258), a randomized phase 3 study. The primary outcome is Change in HbA1c From Baseline at Week 24.

2. Clinical Question

HbA1c is the named primary measure of glycemic control in this trial. The registry reports least-squares mean changes for each arm and between-arm analyses. Those are statistical comparisons of recorded changes; they do not, by themselves, establish what an individual participant experienced or explain why an outcome changed.

The trial also recorded systolic blood pressure and body-weight change as secondary outcomes. These provide additional outcome information, but they are not substitutes for the primary question and should be interpreted with attention to their uncertainty and their secondary status. Safety information is posted separately as counts of serious and other adverse events, along with a death count. Those summaries describe events recorded in each group and do not establish whether an event was caused by the study drug.

3. Trial Design

The registry identifies this as a phase 3, randomized, parallel-group study with quadruple masking. Its official title describes a multi-center, double-blind, placebo-controlled comparison, while the masking field records quadruple masking. The registry does not specify which people were included in that masking designation, so the exact masked roles are not reported in the registry record.

The study compared bexagliflozin tablets, 20 mg, with placebo tablets. Both were to be self-administered once daily for 24 weeks. The registry lists 210 participants as the enrollment and gives the study status as completed. The study start is recorded as 2016-03, and primary completion as 2017-04. Theracos is listed as the sponsor.

The primary outcome was change in HbA1c from baseline at 24 weeks. The analysis population is identified as the intention-to-treat population. The registry also lists secondary outcomes for change in systolic blood pressure and change in body weight among participants with a BMI ≥ 25 Kg/m2. Its record contains 6 outcome measures and 19 analyses, although only 2 secondary outcomes are posted in the information summarized for these results.

The registry record does not report the planned allocation ratio, the original sample-size justification, detailed eligibility criteria, or the full sequence of study visits. It also does not report enough detail to reconstruct the complete protocol or to determine how every analysis in the record relates to each other. These absences limit how much can be concluded about design choices beyond the randomized, parallel-group, masked comparison and the stated treatment period.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: A total of 210 subjects were randomized to be in the bexagliflozin arm or in the placebo arm in a ratio of 2:1.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention-to-Treat Population

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 HbA1c From Baseline at Week 24, measured over 24 weeks.

The primary endpoint is change in HbA1c from baseline at 24 weeks, measured as a percentage of HbA1c. HbA1c is a continuous laboratory measure, so the posted results describe average changes rather than event counts. The registry identifies the estimate as a least-squares mean and the accompanying measure of dispersion as a standard error.

The registry says HbA1c was measured using an HPLC method in laboratories that had completed NGSP Level I laboratory certification and were traceable to the Diabetes Control and Complications Trial reference method. This describes the measurement approach and laboratory standardization noted in the record. The posted outcome is therefore a laboratory measure, not a direct report of symptoms or of an individual participant’s experience.

In general, a continuous endpoint can retain information about the magnitude of change across participants, rather than reducing results to a simple count of people who did or did not cross a threshold. The group mean helps describe the average observed change, while a between-group contrast addresses how the arms differed on average. Neither quantity describes the full distribution of individual outcomes. The registry does not report a threshold for a clinically important change, so the size of the estimate should not be assigned a clinical importance based on this record alone.

7. Results Posted in the Registry

• Bexagliflozin Tablets, 20 mg: mean -0.51 versus -0.10 on control; difference -0.41 (95% CI -0.68 to -0.14), standardized effect d = -0.44; z-test p = 0.002.

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.

Change in Systolic Blood Pressure (SBP) From Baseline at Week 24

• 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:
  • Bexagliflozin Tablets, 20 mg: mean -0.60 versus 1.54 on control; difference -2.14 (95% CI -5.88 to 1.60), standardized effect d = -0.16; z-test p = 0.262.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Bexagliflozin Tablets, 20 mg and the control. The interval gives the range of effects that remain compatible with the data.

Change in Body Weight From Baseline at Week 24 in Subjects With a BMI ≥ 25 Kg/m2

• 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:
  • Bexagliflozin Tablets, 20 mg: mean -1.85 versus -1.06 on control; difference -0.79 (95% CI -2.01 to 0.43), standardized effect d = -0.19; z-test p = 0.206.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Bexagliflozin Tablets, 20 mg and the control. The interval gives the range of effects that remain compatible with the data.

The two posted secondary outcomes are systolic blood-pressure change and body-weight change among participants with a BMI ≥ 25 Kg/m2. Across these outcomes, the reported contrasts point in a negative direction, but the sponsor-reported confidence intervals include zero. Taken together, the secondary results therefore do not provide comparably clear evidence of a between-arm difference in these measures in the posted analyses.

This pattern is distinct from the primary HbA1c result, whose reported sponsor intervals are entirely negative. The outcomes address different measures, so the secondary analyses should be viewed as additional information rather than as confirmation of the primary result. Their estimates do not establish that the primary result necessarily extends to blood pressure or body weight, nor do they resolve questions about the primary analysis.

In general, testing several outcomes or conducting several analyses increases the opportunity to observe apparently notable results by chance. A multiplicity procedure can adjust how evidence is assessed across a family of tests, but whether such a procedure was prespecified or used is not reported in the registry record. The secondary results consequently carry less weight than a prespecified primary outcome for interpreting the study, especially without a reported testing hierarchy or multiplicity adjustment. The record also does not report enough detail to determine whether all secondary analyses planned by the protocol are included in the posted results.

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

• Change in Systolic Blood Pressure (SBP) From Baseline at Week 24
• Change in Body Weight From Baseline at Week 24 in Subjects With a BMI ≥ 25 Kg/m2

9. Statistical Interpretation

10. Statistical Methods Explained

The registry describes the primary endpoint as a least-squares mean change in a continuous measure, with standard errors reported for the arm estimates. It also records sponsor comparisons using mixed models analysis and ANCOVA, with the parameter described as mixed-effects repeated measures. These labels indicate that the reported contrasts come from analyses that account for repeated outcome measurements and use model-based adjustment; however, the complete model specification, covariates, covariance structure, and missing-data assumptions are not reported in the registry record.

What does a least-squares mean represent?

In general, a least-squares mean is a model-based estimate of an arm’s average outcome, typically conditional on the terms included in the analysis model. It is not necessarily the same as a raw arithmetic mean calculated directly from all recorded observations. The registry does not list the model terms in enough detail to explain precisely how the adjustment was made in these analyses. The term should therefore be read as the sponsor’s model-based arm estimate, not as a fully reproducible calculation from the posted summary alone.

What do standard errors describe?

In general, a standard error describes the estimated sampling variability of an estimate. It conveys how much the estimate might vary across repeated samples under the assumptions of the analysis. A standard error is not the spread of individual participant outcomes; it concerns uncertainty in the estimated mean. The arm-level standard errors are posted, but the registry does not provide the individual-level outcome data needed to independently reproduce the fitted models.

Why use repeated-measures models?

In general, a repeated-measures model can use outcome observations collected at multiple time points and represent the relationship among measurements from the same participant. This can help estimate change over time while accounting for within-participant dependence. A model’s results depend on its specification and assumptions. The record names mixed models and repeated measures but does not report the full specification or the handling of missing measurements, so those details cannot be assessed from the registry record.

ANCOVA is a regression approach that can compare groups while adjusting for other variables included in the model. In general, adjustment may improve precision or address baseline differences when the model is appropriately specified. The record identifies ANCOVA analyses but does not state all included covariates. It is therefore not possible to determine from these results alone exactly which adjustments were used or how they relate to the mixed-model entries.

How should the recalculation be understood?

The recalculated comparison is a simple, unadjusted contrast based on the posted arm summary statistics. The posted arm estimates are -0.51 for bexagliflozin and -0.1 for placebo, and the recalculated contrast is -0.41. The recalculation uses approximate standard deviations inferred from the posted standard errors, as noted in the calculation details. That approximation does not recreate a repeated-measures model, account for covariates, or reproduce any participant-level missing-data procedure. It is a useful check on the direction and broad scale of the posted arm summaries, but it is not the sponsor’s model-based analysis.

The recalculated confidence interval and p-value belong to that simple comparison. In general, a p-value describes how compatible the data are with a specified null hypothesis under the analysis assumptions; it is not the probability that the null hypothesis is true. A confidence interval describes a range of estimates reasonably compatible with the data and model at its stated confidence level. These general interpretations do not remove the limitations of an approximate calculation. The sponsor-reported results should be used to describe the trial’s own analyses.

What does the analysis population tell us?

The primary outcome is identified as using the intention-to-treat population. In general, an intention-to-treat approach analyzes participants according to their randomized group, regardless of subsequent treatment adherence, subject to the trial’s stated analysis rules. The label alone does not establish how unavailable measurements, withdrawals, or other missing outcomes were handled. Those procedures are not reported in the registry record.

Finally, the registry lists a superiority analysis for the posted comparisons. In general, a superiority analysis tests whether the groups differ in the specified direction or whether the data support a difference under its model. The registry does not provide a complete statistical analysis plan, so it is not possible to verify prespecified model details, testing order, or all decision rules from the record alone.

11. Confidence Intervals and Precision

The primary sponsor analyses report 95% confidence intervals for the between-arm HbA1c contrast. The mixed-model result gives an estimate of -0.41 with an interval from -0.66 to -0.16; the ANCOVA result gives -0.41 with an interval from -0.68 to -0.15. Both intervals are entirely on the negative side of zero. This indicates that the reported model-based estimates are compatible with a range of negative contrasts under the respective analyses, rather than identifying one exact value.

In general, a confidence interval expresses uncertainty about an estimated quantity under the assumptions of the analysis. A wider interval represents less precision than a narrower interval, all else equal. An interval should not be read as a range containing every participant’s outcome or as a guarantee that the true effect lies within it. It is tied to the model and analysis that produced it.

The registry includes other primary analyses with different point estimates, including -0.55 and -0.40. Their reported 95% intervals are also negative. The differences among these entries underscore that the reported result depends in part on the analysis used. The record does not fully distinguish the purpose and specification of each entry, which limits a detailed explanation of why the estimates differ.

The secondary-outcome intervals are less decisive: their reported intervals cross zero. In general, when an interval includes zero for a difference, the data and model are compatible with no between-group difference as well as with values on either side of zero. That is not proof that the groups are identical; it reflects uncertainty at the displayed precision. The registry’s reported intervals should be interpreted in the context of the analyses and the secondary status of those outcomes.

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 sample-size display gives planning illustrations in participants per arm for these continuous-outcome comparisons; it does not report total event counts. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each illustration uses a stated fraction of the observed effect as the effect to detect. The original trial design assumptions and sample-size justification are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation estimates would be needed to have an 80% chance of detecting the specified effect under those assumptions, if that effect were the true effect. The 90% power column uses the same idea with a higher target chance of detecting that effect. These are planning targets, not descriptions of what happened in the trial. The listed values are participants per arm, not the combined number across both arms.

The rows for 100%, 75%, and 50% of the observed effect illustrate how the planning count changes as the target effect changes. Smaller effects require more participants to detect with the same planned power, because they are harder to distinguish from sampling variability. The display uses the observed estimate as its reference point, so its values depend on that estimate and should not be treated as a design rationale or a guarantee of future performance. This is not observed power.

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 = -0.410; pooled standard deviation σ = 0.942.
2. n = 2 × (1.96 + 0.842)² × 0.942² / -0.410² = 82.8, rounded up to 83 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 <- -0.41                      # observed difference in means
sigma <- 0.941849327704576                     # 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(2715258)
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 = 83 (80% power), 111 (90% power)
 75% of observed effect: n per arm = 148 (80% power), 198 (90% power)
 50% of observed effect: n per arm = 332 (80% power), 444 (90% power)
power.t.test: 83.8 
Simulated power at n = 83 per arm: 0.797 (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 registry’s sponsor analyses are model-based: the primary outcome is reported with mixed models analysis and ANCOVA, and the parameter is described as mixed-effects repeated measures. The reported primary estimates include -0.41, with 95% confidence intervals of -0.66 to -0.16 and -0.68 to -0.15. Other listed analyses report -0.55 and -0.40, each with a 95% interval entirely below zero. The registry does not fully explain the differences among the four analysis entries.

The simple recalculation uses the posted arm-level summaries and yields a contrast of -0.41 with an unadjusted p-value of 0.002. It is not the same procedure as fitting the sponsor’s repeated-measures models. The recalculation uses approximate variability derived from the posted standard errors and does not include the model structure or adjustments that may have been used in the sponsor analyses. Its confidence limits therefore need not match the sponsor’s.

The recalculation is best understood as a transparent, approximate check on the posted summary statistics. The sponsor’s analyses describe the trial’s reported statistical results. The registry does not flag a discrepancy between the recalculation and the posted result, but that does not make the methods identical or establish that the simple calculation reproduces the trial analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 6 outcome measures and 19 analyses, while the results summarized for the trial include the primary HbA1c outcome and 2 secondary outcomes. Multiple outcomes and multiple analyses create a multiplicity question: when many statistical tests are considered, the chance of at least one apparently unusual result can increase. In general, a prespecified testing hierarchy or adjustment can define how results across tests should be interpreted. Whether the study used such an approach is not reported in the registry record.

The registry labels the primary comparisons as superiority analyses and reports results from mixed models analysis and ANCOVA. It does not report a complete statistical analysis plan, a primary testing sequence, any family-wise error procedure, or whether analyses were considered confirmatory or exploratory beyond the outcome labels. Without those details, the reported p-values should not be read as if the handling of all tests were known. The primary endpoint remains the clearest designated outcome, but the record does not establish how inference was coordinated across every listed analysis.

The primary outcome population is described as intention-to-treat. The registry record does not report the number of missing outcome measurements at each time point, participant-level reasons for missingness, imputation procedures, or sensitivity analyses for missing data. It also does not detail whether participants were analyzed according to assigned treatment in every reported model or how departures from treatment were handled. The intention-to-treat label alone does not answer these questions.

In general, repeated-measures models can use available outcome measurements, but their interpretation depends on assumptions about the data and missingness. The registry’s method labels do not specify those assumptions. It is therefore not possible to assess from the posted record how robust the primary estimates are to different missing-data mechanisms or analytic choices. These omissions limit reproducibility and make it important to distinguish the posted estimates from a complete account of the trial’s statistical plan.

16. 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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cardiomyopathy (1.4%), tenosynovitis (0.7%).

The registry reports serious adverse events in 1 of 138 participants in the bexagliflozin arm and 1 of 69 participants in the placebo arm. It lists no deaths in the bexagliflozin arm and 1 death in the placebo arm. The record also reports other adverse events in 41 of 138 participants receiving bexagliflozin and 25 of 69 receiving placebo. These counts describe the groups as recorded; they do not, by themselves, establish a causal relationship between an event and assigned treatment.

The listed serious-event terms are cardiomyopathy, recorded as 1.4%, and tenosynovitis, recorded as 0.7%. The record does not provide enough detail in these summaries to assess event timing, severity beyond the serious-event designation, clinical course, or attribution. The displayed counts and terms should therefore be read as descriptive safety information, not as a complete safety profile or a comparative risk estimate.

In general, adverse-event comparisons require attention to participant counts, follow-up, event definitions, and the way repeated events are counted. The registry summaries give counts of participants affected and the denominators at risk, but do not answer all of those questions.

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

The main statistical finding is that the registry’s reported analyses estimate a more negative average HbA1c change in the bexagliflozin group than in the placebo group over 24 weeks. The sponsor-reported primary contrasts are negative, and their 95% confidence intervals exclude zero. The simple recalculation from posted arm summaries points in the same direction, although it does not reproduce the sponsor’s modeling.

This result is an average comparison between randomized groups. It does not show that every participant had the same change, that an individual participant would experience a particular outcome, or that the estimated contrast has a specific clinical importance. The registry does not report a clinically important threshold for the endpoint. Statistical evidence about a group average should not be equated with a prediction for an individual.

The result also does not establish what would happen beyond the 24-week timeframe, in populations with characteristics not described in the record, or under conditions different from the trial’s once-daily regimen and placebo comparison. The registry’s primary analysis population is intention-to-treat, but detailed missing-data procedures and full model specifications are not reported. Those limits matter when assessing how robustly the estimate represents the randomized comparison.

The secondary outcomes and safety counts add context but do not settle questions about the primary measure’s clinical significance or longer-term consequences. The secondary intervals include zero, and adverse-event summaries alone do not establish causation or comparative safety. The record supports a description of the posted statistical results and their uncertainty; it does not support broader claims beyond those outcomes, analyses, participants, and follow-up recorded in the registry.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical result should be read through both its estimate and its uncertainty. A difference between arm means provides a measure of the average contrast, while a confidence interval communicates how much uncertainty remains under the analysis. Reporting both is more informative than treating a p-value as a complete summary of evidence.

It also demonstrates that an arm-summary recalculation and a trial’s own model-based analysis answer closely related but not identical statistical questions. The simple recalculation uses posted group summaries; the sponsor analyses use methods identified as repeated-measures mixed models and ANCOVA. Their estimates are broadly similar in direction, but model details and missing-data handling are not fully described in the registry. Transparent distinctions among analyses help readers avoid treating an approximation as a reproduction of the trial analysis.

The secondary results show the importance of separating the primary endpoint from additional outcomes. The registry records multiple outcome measures and analyses, but does not report a multiplicity plan. Without that context, secondary p-values and estimates require caution and should not be given the same interpretive weight as a designated primary outcome.

Finally, the planning illustrations demonstrate how sample-size requirements depend on the effect a study aims to detect and on stated assumptions. They are not evidence that a trial was adequately powered for every outcome, nor do they reveal the original design rationale. Together, these features make the record useful for understanding the reported comparison while also showing why protocol-level statistical details matter for a complete appraisal.

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 registry record leaves several trial-specific questions unresolved. It gives an enrollment of 210, while the primary arm results list 138 participants for bexagliflozin and 69 for placebo. The record does not explain the relationship between enrollment and the analysis counts, including whether they reflect different analysis sets or outcome availability. The intention-to-treat label does not resolve this on its own.

Four primary analysis entries are posted, including estimates of -0.41, -0.55, and -0.40. Although each reported interval is negative, the registry does not fully describe why the analyses differ, which model specification corresponds to each entry, or whether they represent distinct populations, time-point procedures, or sensitivity analyses. This makes it difficult to identify a single model as the definitive result from the registry record alone.

Detailed baseline characteristics, eligibility criteria, adherence information, participant flow, missingness by visit, and full statistical model specifications are not reported in the registry record. The record also does not report the original sample-size justification, multiplicity approach, or a complete protocol-level analysis plan. These details would help assess representativeness, reproducibility, and the role of each analysis.

Follow-up for the primary endpoint is 24 weeks. The registry results therefore do not describe outcomes beyond that period. Safety summaries are limited to the posted counts and terms and do not give a full account of event timing, duration, attribution, or recurrence. These limitations restrict interpretation to the recorded outcomes and available statistical summaries.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Safety and Efficacy of Bexagliflozin as Monotherapy in Patients With Type 2 Diabetes — NCT02715258.
• Registry record last updated 2021-06-28; status last verified 2021-06.
• Results first posted 2021-04-28.
• 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

NCT02715258 is a completed phase 3 randomized, quadruple-masked, parallel-group comparison of bexagliflozin tablets, 20 mg, with placebo in participants with type 2 diabetes and inadequate glycemic control. Both regimens were once daily for 24 weeks. Theracos is listed as sponsor; the recorded study start is 2016-03 and primary completion is 2017-04.

The primary endpoint was change in HbA1c from baseline at 24 weeks in the intention-to-treat population. Sponsor analyses report negative between-arm estimates with 95% confidence intervals below zero. The posted secondary outcomes concern systolic blood pressure and body weight; their reported intervals include zero. Safety summaries list serious and other adverse events and deaths by arm.

The registry does not fully report the statistical plan, multiplicity handling, missing-data procedures, or distinctions among all primary analyses. Results should be interpreted as the posted estimates for this trial and timeframe, with those reporting limits in mind.