A Study to Test the Efficacy, Safety and Tolerability of Romosozumab Treatment in Postmenopausal Chinese Women With Osteoporosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Test the Efficacy, Safety and Tolerability of Romosozumab Treatment in Postmenopausal Chinese Women With Osteoporosis (NCT05067335), a randomized phase 3 study. The primary outcome is Percent Change From Baseline in Bone Mineral Density (BMD) at the Lumbar Spine to the End of Double-Blind Period.

2. Clinical Question

The study also assessed BMD at the total hip and femoral neck during the double-blind period, as well as BMD outcomes at Month 12. These measurements broaden the description of the observed pattern across skeletal sites and follow-up periods; they do not replace the designated primary question. The Month 12 comparison has a different treatment context because the placebo-assigned group had started romosozumab during the open-label period. It therefore should not be read as a continuing placebo-controlled comparison.

3. Trial Design

The registry describes a Phase 3, randomized, multicenter, double-blind, placebo-controlled, parallel-group study. Masking is recorded as quadruple. The registry does not identify the specific roles covered by that masking designation, so those details are not reported in the registry record. The listed enrollment is 327, and the study status is completed.

The study compares romosozumab with placebo during a double-blind placebo-controlled period. Participants randomized to romosozumab received romosozumab during all treatment periods. Participants randomized to placebo received placebo during the double-blind period and romosozumab during the open-label treatment period. The primary endpoint timeframe runs from baseline on Day 1 to the end of the double-blind period at Month 6. Some secondary BMD outcomes are listed at Month 12.

The registry lists the study start as 2021-10-21 and primary completion as 2023-11-09. UCB Biopharma SRL is identified as the sponsor. The record does not report the original sample-size justification, the planned allocation ratio, or the trial’s original assumptions for its design calculations. Those omissions prevent a reconstruction of the design rationale from the registry record alone. The planning illustrations associated with posted results use a separate set of stated assumptions and should not be confused with the original trial design.

Randomization supports a between-group comparison during the controlled period, while blinding is intended in general to limit the influence of expectations on care and assessment. These are general design considerations; the record does not detail implementation procedures or the roles masked. The open-label follow-up changes the interpretation of later comparisons because both treatment histories include romosozumab.

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: Participant flow refers to the Randomized Set.

Double-blind Period: Day 1 - Month 6

Reasons for not completing this period:

Open-label Period: Month 7 - 12

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Full Analysis Set (FAS) consisted of all randomized study participants who received at least 1 dose of investigational medicinal product (IMP) and provided at least 1 Baseline and post-Baseline BMD measurement.

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: Baseline Characteristics refer to the RS which consisted of study participants who were randomized into this study.

6. Primary Endpoint

The registry lists the primary outcome as Percent Change From Baseline in Bone Mineral Density (BMD) at the Lumbar Spine to the End of Double-Blind Period, measured over From Baseline (Day 1) to the end of the Double-blind Period (Month 6).

The primary endpoint is percent change from baseline in bone mineral density at the lumbar spine, measured by dual-energy x-ray absorptiometry (DXA) at the end of the double-blind period, Month 6. It describes the change in a continuous measurement relative to baseline, expressed as a percentage. The registry reports a least-squares mean for each group and an adjusted difference between those means.

In general, BMD is used as a quantitative measure of bone density, and specifying a measurement site and timepoint makes the outcome interpretable and reproducible. A baseline-relative endpoint focuses the comparison on change during the stated study interval rather than only on the final measurement. The lumbar spine is the prespecified primary site in this registry record; total hip and femoral neck measurements are listed as secondary outcomes.

The primary analysis population is the Full Analysis Set. It includes randomized participants who received at least one dose of investigational medicinal product and had both a baseline and a post-baseline BMD measurement. The registry describes missing-data handling for the primary endpoint: non-COVID-related missing post-baseline BMD was imputed using last observation carried forward, while specified COVID-related missing or out-of-window measurements were imputed using multiple imputation under a missing-at-random approach. The implications of these choices depend on assumptions about the unobserved measurements, which cannot be verified from summary results alone.

7. Results Posted in the Registry

• Romosozumab: mean 9.81 versus 0.44 on control; difference 9.37 (95% CI 8.30 to 10.44), standardized effect d = 2.01; z-test p = <0.001.

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 From Baseline in Bone Mineral Density at the Total Hip to the End of Double-Blind Period

• 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:
  • Romosozumab: mean 2.93 versus 0.07 on control; difference 2.86 (95% CI 2.17 to 3.55), standardized effect d = 0.96; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Romosozumab than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percent Change From Baseline in Bone Mineral Density at the Femoral Neck to the End of the Double-Blind Period

• 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:
  • Romosozumab: mean 3.33 versus -0.15 on control; difference 3.48 (95% CI 2.43 to 4.53), standardized effect d = 0.73; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Romosozumab than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percent Change From Baseline in Bone Mineral Density at the Lumbar Spine at Month 12

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Romosozumab/Romosozumab Overall: mean 13.04 versus 8.80 on control; difference 4.24 (95% CI 2.91 to 5.57), standardized effect d = 0.75; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Romosozumab/Romosozumab Overall than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary BMD results follow a broadly consistent descriptive pattern across the reported Month 6 sites: the posted romosozumab group estimates are higher than the placebo estimates, and the reported ANCOVA contrasts are positive. The registry also presents a Month 12 lumbar-spine comparison after the placebo-assigned group entered open-label romosozumab treatment. That later comparison reflects treatment histories rather than a continuing placebo-controlled contrast.

The secondary outcomes relate to the primary result by examining other measurement sites and a later timepoint. Agreement in direction can add context to the overall pattern, but it does not independently establish the primary result or remove the need to consider the separate analysis populations, timeframes, and methods. The record lists five secondary outcomes, while detailed outcome summaries and analyses are not reported in the registry record for every listed secondary outcome.

Secondary findings generally carry less confirmatory weight when many outcomes are tested and no multiplicity procedure is reported. Each test adds another opportunity for a result to appear unusual by chance under its null hypothesis. The registry does not report an adjustment for multiplicity or a prespecified hierarchy for interpreting secondary endpoints. Thus, secondary p-values, where reported, should be treated as descriptive or supportive rather than as equivalent to a separately controlled primary claim.

The Month 12 result deserves particular caution in interpretation: placebo participants received romosozumab during the open-label period, so the two group labels encode different treatment sequences. It cannot be interpreted as a direct Month 12 comparison of ongoing romosozumab against ongoing placebo. The registry record does not report detailed inferential analyses for all listed secondary outcomes.

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

• Percent Change From Baseline in Bone Mineral Density at the Total Hip to the End of Double-Blind Period
• Percent Change From Baseline in Bone Mineral Density at the Femoral Neck to the End of the Double-Blind Period
• Percent Change From Baseline in Bone Mineral Density at the Lumbar Spine at Month 12
• Percent Change From Baseline in Bone Mineral Density at the Total Hip at Month 12
• Percent Change From Baseline in Bone Mineral Density at the Femoral Neck at Month 12

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is continuous: percent change from baseline in lumbar-spine BMD. The registry reports least-squares means (LSMs), which are model-based group means, and standard errors for the group summaries. In general, a standard error describes uncertainty in an estimated mean; it is not the same quantity as the spread of individual participant measurements. The posted LSMs are 0.44 for placebo and 9.81 for romosozumab, with standard errors of 0.442 and 0.323, respectively.

What does ANCOVA contribute?

The registry reports an analysis of covariance (ANCOVA) for the primary comparison. In general, ANCOVA estimates a group contrast while accounting for covariates included in the model. This can differ from a direct comparison of group summaries. The registry reports the method and the resulting difference in LSMs, but the covariates and further model specifications are not reported in the registry record. It is therefore not possible to assess from these summaries exactly how adjustment affected the estimate.

The reported contrast is romosozumab minus placebo. Its estimate is 9.37, with a 95% confidence interval from 8.34 to 10.39 and p < 0.001. The p-value describes how compatible the observed test statistic is with a null hypothesis of no difference, under the model and its assumptions. In general, it is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect.

Why do the recalculated values differ slightly?

The posted recalculation uses the group summaries and is explicitly an unadjusted calculation. For the primary endpoint, it gives a difference of 9.37 and an interval from 8.297 to 10.44. Its inputs include standard errors and an approximate pooled standard deviation. The sponsor’s reported ANCOVA result uses the trial’s analysis method. Small differences in interval endpoints can occur when an unadjusted calculation and a model-based analysis use different information or assumptions. The reported analysis is the trial’s principal result; the recalculation is a supplementary statistical check, not a replacement for it.

What is the role of missing-data methods?

For the primary endpoint, non-COVID-related missing post-baseline BMD measurements were imputed using last observation carried forward (LOCF). In general, LOCF carries a participant’s last observed value forward to represent a later missing measurement. That approach can be sensitive to whether the last value is a reasonable representation of the missing value. For measurements missing or outside the allowed window due to COVID-19, the record says values were set to missing and imputed by multiple imputation under a missing-at-random approach. The out-of-window criterion is described as exceeding 70 days since the previous investigational medicinal product. In general, multiple imputation represents uncertainty about missing values by using a model to generate plausible values; a missing-at-random assumption means missingness is explained by observed information included in the model. The imputation model details and sensitivity analyses are not reported in the registry record.

The primary analysis population is the Full Analysis Set: randomized participants who received at least one dose of investigational medicinal product and contributed baseline and post-baseline BMD measurements. This population definition is narrower than all randomized participants because it includes dose and measurement requirements. The number of participants included in the endpoint summaries is listed by group, but the registry does not give a full account of how many randomized participants were excluded from this analysis or why.

How should the secondary results be viewed?

The registry reports ANCOVA comparisons for the primary endpoint and for the two Month 6 secondary endpoints at the total hip and femoral neck. The Month 12 lumbar-spine endpoint has group summaries and a posted unadjusted recalculation but no reported ANCOVA analysis in the available record. Further details on the analysis methods for the other listed secondary endpoints are not reported in the registry record.

In general, evaluating multiple outcomes creates a multiplicity issue: with more statistical tests, the chance of at least one apparently small p-value can increase if each test is interpreted independently. The record does not report a multiplicity adjustment or a hierarchical testing procedure. Consequently, secondary p-values should be treated as less confirmatory than the designated primary analysis. No interim analyses or complete statistical analysis plan are reported in the registry record.

These methodological distinctions matter because the primary result, recalculated summaries, and secondary outcomes answer related but non-identical questions. The registry’s reported model, its population definition, and its missing-data assumptions define the scope of the primary inference. General statistical explanations in this discussion describe how these methods usually work; they do not add unreported details about the trial’s implementation.

11. Confidence Intervals and Precision

The sponsor-reported primary difference is 9.37 percentage points, with a 95% confidence interval from 8.34 to 10.39. The interval describes a range of values compatible with the data under the ANCOVA model and its assumptions. In general, a 95% confidence interval is a procedure that would contain the target quantity in 95% of repeated samples under the model conditions; it is not a statement that there is a 95% probability that this particular fixed parameter lies inside this particular interval.

The primary interval lies above zero, which aligns with the reported p < 0.001 for a test of no difference. Its width also conveys uncertainty: the observed estimate is not known with arbitrary precision, and values across the interval remain compatible with the analysis. The interval does not capture every source of uncertainty, such as uncertainty about model choice, missing-data assumptions, or generalizability beyond the enrolled population.

The recalculated interval runs from 8.297 to 10.44. It is based on a simple unadjusted calculation from posted summary statistics, rather than the reported ANCOVA model. Its close numerical relationship to the sponsor-reported interval is reassuring as a consistency check, but the two intervals are not interchangeable. The registry’s reported ANCOVA interval is the relevant interval for the trial’s stated analysis.

Standard errors are also listed for the group LSMs. In general, a standard error reflects the sampling precision of an estimate, while a confidence interval communicates uncertainty on the scale of the estimated effect. Neither one describes the spread of individual participant responses. The registry does not report a clinical-importance threshold, so the interval’s statistical precision cannot by itself determine the practical importance of the estimated difference.

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 illustrations are expressed as participants per arm, not total events. They use two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set in turn to the stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record, so these illustrations should not be read as a reconstruction of the original design.

The 80% power column gives the number of participants per arm that the illustration estimates would be needed for an 80% chance of detecting the specified effect under those assumptions, if that effect were the true effect and the model assumptions held. The 90% power column expresses the corresponding planning quantity for a 90% chance. These are planning concepts, not guarantees that a future study will achieve the stated power.

Each row uses a fraction of the observed effect: the full observed effect, 75% of that effect, or 50% of that effect. The smaller target effects require more participants per arm in the listed illustrations. For example, the primary endpoint’s illustration rises from 4 participants per arm for 80% power at 100% of the observed effect to 16 at 50% of the observed effect; the corresponding 90% figures are 6 and 21. These values depend strongly on the observed estimate and the planning assumptions, so they are not universal sample-size recommendations.

This is not observed power. It is a hypothetical planning calculation based on the observed effect and stated assumptions, and it does not show the probability that the completed trial’s result is correct. The posted interval and the trial’s reported analysis are more informative for interpreting the uncertainty in the observed estimate.

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 = 9.370; pooled standard deviation σ = 4.655.
2. n = 2 × (1.96 + 0.842)² × 4.655² / 9.370² = 3.9, rounded up to 4 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 <- 9.37                      # observed difference in means
sigma <- 4.65528950751016                     # 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(5067335)
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 = 4 (80% power), 6 (90% power)
 75% of observed effect: n per arm = 7 (80% power), 10 (90% power)
 50% of observed effect: n per arm = 16 (80% power), 21 (90% power)
power.t.test: 5 
Simulated power at n = 4 per arm: 0.658 (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 reported primary analysis is an ANCOVA comparison of least-squares means. It gives a romosozumab-minus-placebo difference of 9.37, a 95% confidence interval from 8.34 to 10.39, and p < 0.001. This is the sponsor-reported inferential result and should be used as the primary reference when interpreting the trial’s endpoint.

The recalculated result uses the posted group summaries for a simple unadjusted comparison. It gives the same difference estimate, 9.37, and a confidence interval from 8.297 to 10.44. The estimates and intervals are close, but the calculation does not reproduce ANCOVA and uses approximate summary-statistic inputs. It serves as a check on the scale and direction of the reported result, not as a substitute for the trial analysis.

The registry indicates no discrepancy between the recalculation and reported result. The reported ANCOVA has the more direct relevance to the trial’s specified primary analysis because it is the analysis method identified in the registry. The registry does not report ANCOVA covariates or complete model details, so those cannot be independently reconstructed from the posted summaries.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary endpoint and lists secondary BMD outcomes at the total hip, femoral neck, and lumbar spine across the double-blind and Month 12 timeframes. It reports the primary ANCOVA analysis and ANCOVA results for the Month 6 total-hip and femoral-neck outcomes. The record states there are eight outcome measures and three analyses. A full multiplicity strategy, such as a prespecified testing hierarchy or adjustment of significance thresholds, is not reported in the registry record.

In general, multiplicity matters when several outcomes or comparisons are tested: interpreting every p-value at the same nominal threshold can raise the chance of at least one false-positive result across the collection of tests. Without a stated adjustment or hierarchy, the primary endpoint should be distinguished from secondary analyses, and unadjusted secondary findings should not be treated as if they had the same confirmatory status as the primary result. The registry does not report interim analyses or enough of the statistical analysis plan to establish how all planned tests were intended to be interpreted.

Missing data are addressed specifically for the primary BMD endpoint. The registry says missing post-baseline measurements not due to COVID-19 were imputed with last observation carried forward. Measurements missing or outside the defined window due to COVID-19 were set to missing and imputed through multiple imputation under a missing-at-random approach. The record describes the out-of-window condition as exceeding 70 days since the previous investigational medicinal product. These methods rely on assumptions about values that were not observed. The specific imputation models, diagnostics, and sensitivity analyses are not reported in the registry record.

The Full Analysis Set consists of randomized participants who received at least one dose and had at least one baseline and one post-baseline BMD measurement. This definition is explicit, but details on exclusions and the number of participants with missing endpoint data are not reported in the registry record. Missing-data methods for secondary endpoints are also not reported in the registry record. These limitations constrain assessment of how robust the estimates are to alternative assumptions about missing measurements.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: appendicitis (1.0%), upper respiratory tract infection (1.0%), femoral neck fracture (1.0%), intervertebral disc protrusion (1.0%), adenocarcinoma gastric (1.0%).

The registry presents serious adverse event counts by listed safety group, alongside the number at risk. In the placebo group, 5 serious adverse events are listed among 109 at risk; in the romosozumab group, 8 are listed among 218 at risk. For the placebo/romosozumab group, the figures are 9 among 98 at risk, and for the romosozumab/romosozumab overall group, 25 among 218 at risk. These are counts as recorded, not estimates of comparative risk adjusted for exposure or follow-up.

The group labels reflect different treatment sequences and reporting categories. The placebo-assigned participants later received romosozumab, and the registry also gives overall sequence-group summaries. The record does not provide enough detail to treat all group counts as mutually exclusive or to combine them. No deaths are listed in the placebo, romosozumab, or placebo/romosozumab groups; one death is listed in the romosozumab/romosozumab overall group.

The registry lists several serious adverse event terms at 1.0% each, including appendicitis, upper respiratory tract infection, femoral neck fracture, intervertebral disc protrusion, adenocarcinoma gastric, and cerebrovascular insufficiency. A term’s listing does not establish whether it was caused by treatment. These counts alone do not support a comprehensive comparison of safety, and the registry record does not report an adjusted safety analysis in the information.

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

The primary result estimates the difference in mean percent change in lumbar-spine BMD between randomized groups at the end of the double-blind period. The reported ANCOVA estimate is positive, and its 95% confidence interval lies above zero. This is evidence of a difference under the trial’s reported model and analysis population; it is not a guarantee about the result for every participant or every setting.

The result does not establish the size of any individual participant’s change. Group means summarize central tendencies, and standard errors describe uncertainty in those estimated means rather than participant-to-participant variability. The registry record does not provide the full distribution of individual changes in the information used for this analysis.

The endpoint is a measurement of BMD, not a direct report of fractures, symptoms, function, or quality of life. Although the registry includes serious adverse event counts, those counts do not by themselves provide a complete benefit-risk assessment. Clinical importance thresholds for the BMD change are not reported in the registry record, and the statistical result should not be taken as a substitute for such context.

The controlled comparison concerns the double-blind period through Month 6. Later results should be interpreted in light of the placebo group’s transition to romosozumab during the open-label period. The Month 12 contrast therefore does not represent continued placebo versus romosozumab. The study population is postmenopausal Chinese women with osteoporosis, so the posted results do not automatically establish the same estimates in other populations.

Finally, interpretation depends on the reported ANCOVA model, the Full Analysis Set definition, and the specified imputation approaches. Details of the model covariates, the full analysis plan, and sensitivity analyses are not reported in the registry record. The result is best understood as an estimate for a defined endpoint, time period, population, and analysis—not as a conclusion about unmeasured outcomes or treatment beyond the studied conditions.

18. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation depends on more than a p-value. The primary result includes a group contrast, a confidence interval, an analysis method, an endpoint definition, and a stated analysis population. Reading these elements together gives a clearer account of what was estimated and how much uncertainty remains than relying on a significance label alone.

It also shows the distinction between a model-based trial analysis and an unadjusted recalculation from summary statistics. The estimates are numerically similar, but the ANCOVA result is the reported primary analysis. A recalculation can serve as a check on the reported scale and direction, while still differing in method and inferential role. Statistical transparency requires keeping those results distinct rather than presenting them as interchangeable.

The study’s missing-data approach is another important part of the interpretation. LOCF and multiple imputation under a missing-at-random assumption make different choices about unobserved measurements. Because the registry does not report detailed imputation models or sensitivity analyses, readers cannot fully assess how alternative missing-data assumptions might affect the estimates.

Finally, the transition from placebo to open-label romosozumab changes the meaning of later comparisons, and the lack of a reported multiplicity strategy limits the confirmatory weight of multiple secondary findings. These are not merely technical details: treatment sequence, endpoint timing, and multiplicity determine what a contrast can legitimately address. The record is most informative when the primary controlled-period comparison is kept separate from later and secondary analyses.

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 does not report the original sample-size justification or the assumptions used to plan enrollment. The planning illustrations based on observed effects are not a substitute for that information. ANCOVA is identified as the primary method, but its covariates and other model specifications are not reported, limiting independent assessment of the adjusted estimate.

Missing-data handling is described for the primary endpoint, but the imputation model details and sensitivity analyses are not reported in the registry record. The Full Analysis Set definition requires treatment exposure and BMD measurements at baseline and after baseline; the record does not provide a complete account of exclusions or missing endpoint data. Methods for missing secondary outcomes are also not reported.

The record does not report a multiplicity adjustment or a hierarchical testing strategy for the multiple secondary outcomes. Some listed secondary outcomes do not have detailed results or inferential analyses in the information reported. The Month 12 comparison follows a switch from placebo to open-label romosozumab, so it is not a continuing randomized placebo-controlled contrast.

Safety results are presented as counts by listed categories, without a full adjusted comparison or enough information to combine categories. The registry record does not report a clinical-importance threshold for the primary BMD change. These limitations mean that the posted summaries support interpretation of the specified comparisons, but do not resolve broader questions about unreported analyses, other outcomes, or generalizability beyond the study population.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Test the Efficacy, Safety and Tolerability of Romosozumab Treatment in Postmenopausal Chinese Women With Osteoporosis — NCT05067335.
• Registry record last updated 2026-09-28; status last verified 2026-08.
• Results first posted 2024-12-05.
• 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

The registry describes a completed Phase 3 randomized, multicenter, double-blind, placebo-controlled, parallel-group study of romosozumab in postmenopausal Chinese women with osteoporosis. The listed enrollment is 327. The primary endpoint is percent change from baseline in lumbar-spine BMD at Month 6 during the double-blind period.

The sponsor-reported ANCOVA difference in least-squares means, romosozumab minus placebo, is 9.37, with a 95% confidence interval from 8.34 to 10.39 and p < 0.001. The posted recalculation is unadjusted and gives a similar estimate, but the ANCOVA is the registry’s reported primary analysis. Secondary outcomes cover other BMD sites and Month 12; the later comparison follows the placebo group’s switch to romosozumab. Missing-data methods are described for the primary endpoint, while details on multiplicity, model specifications, and several other analyses are not reported in the registry record.