This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Insulin Resistance, Polycystic Ovary Syndrome, and Bone Research Study (NCT00640224), a randomized phase 4 study. The primary outcome is Peripheral Insulin Sensitivity at Baseline and 6 Months..
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00640224 |
| Phase | Phase 4 |
| Condition | Polycystic Ovary Syndrome |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Enrollment | 65 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Silva Arslanian |
| Results posted | Yes |
2. Clinical Question
Population
The treatment groups comprised treatment-naive, overweight adolescent females with PCOS. The registry record also lists overweight/obese adolescents without PCOS and lean girls without PCOS as potential comparison groups, but reports that no participants were enrolled in either group.
Intervention
One randomized treatment was rosiglitazone. The record identifies it as a drug and describes its use in treatment-naive overweight adolescent females with PCOS.
Comparator
The active comparator was drospirenone/ethinyl estradiol, also described as a drug and used in treatment-naive overweight adolescent females with PCOS. The registry record does not describe a placebo group.
Primary question
The study purpose was to investigate the effects of the treatments on insulin sensitivity and androgen levels, as well as inflammatory markers, vascular markers and bone development. The registered primary outcome was peripheral insulin sensitivity at baseline and 6 months. The registry record does not report the outcome values or specify the exact statistical contrast used to answer the primary question.
This is an active-treatment comparison in adolescents with PCOS, not a comparison between treatment and no treatment. The record also describes two non-PCOS groups as developmental or cardiovascular-marker controls. Because no participants were enrolled in those arms, they cannot serve as observed comparison groups in the posted trial results.
The broad clinical question is therefore whether the registered measures, especially peripheral insulin sensitivity, differed over the assessment period between the two treatment groups. That framing follows the stated purpose and outcome timing; the record does not say whether the primary analysis compared follow-up values, changes from baseline, or another contrast. It also does not report a prespecified hypothesis about which treatment would produce a larger value.
3. Trial Design
The registry record identifies the study as randomized, quadruple-masked and parallel, and lists it as Phase 4 and completed. The enrollment field reports 65. The record gives a start date of 2005-03 and a primary completion date of 2015-07. These registry entries describe the study's design and timeline, but do not establish how participants were recruited, how randomization was generated, or how treatment assignment was concealed; those details are not reported in the registry record.
The parallel design means the treatment groups are described as separate groups rather than as a crossover in which participants receive both treatments. Randomization is intended, in general, to make treatment groups comparable on average at the start of a study. It does not guarantee that every measured or unmeasured characteristic is balanced in a particular trial. The registry record does not describe baseline balance or report a randomization procedure.
The masking field says quadruple, but the record does not identify which roles were masked, how masking was maintained, or whether it was assessed. It is therefore possible to report the registry's masking designation, but not to establish from this record which participants or study personnel were unaware of treatment assignment.
The treatment descriptions identify rosiglitazone and drospirenone/ethinyl estradiol as the active treatments. The record does not report dose, dosing schedule, treatment adherence, permitted co-interventions, or reasons for stopping treatment. The outcome timeframe specifies baseline and 6 months for the primary measure; longer-term follow-up is not reported in the registry record.
Although the record describes non-PCOS groups that could have provided context for developmental or cardiovascular-marker comparisons, it explicitly says no participants were enrolled in those arms. The active treatment groups are therefore the only enrolled arms described in the safety data. The registry record does not give a participant flow account or explain how the enrollment figure relates to the safety denominators.
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: 9 did not meet diagnostic criteria, 3 were on psychiatric meds,1 was diagnosed with diabetes,1 had venous access issues, and 5 did not complete baseline evaluations. No participants were enrolled in the "Overweight/Obese without PCOS" or "Lean without PCOS" Arms/Groups.
Overall Study
| Stage | Rosiglitazone | Drospirenone/Ethinyl Estradiol | Overweight/Obese Without PCOS | Lean Without PCOS |
|---|---|---|---|---|
| STARTED | 23 | 23 | 0 | 0 |
| COMPLETED | 17 | 20 | 0 | 0 |
| NOT COMPLETED | 6 | 3 | 0 | 0 |
Reasons for not completing this period:
| Reason | Rosiglitazone | Drospirenone/Ethinyl Estradiol | Overweight/Obese Without PCOS | Lean Without PCOS |
|---|---|---|---|---|
| Lost to Follow-up | 5 | 2 | 0 | 0 |
| Withdrawal by Subject | 1 | 0 | 0 | 0 |
| Headaches | 0 | 1 | 0 | 0 |
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.
| Characteristic | Unit | Rosiglitazone | Drospirenone/Ethinyl Estradiol | Overweight/Obese Without PCOS | Lean Without PCOS | Total |
|---|---|---|---|---|---|---|
| Participants analysed | 17 | 20 | 0 | 0 | 37 | |
| Age, Categorical: Age – <=18 years | Participants | 16 | 18 | — | — | 34 |
| Age, Categorical: Age – Between 18 and 65 years | Participants | 1 | 2 | — | — | 3 |
| Age, Categorical: Age – >=65 years | Participants | 0 | 0 | — | — | 0 |
| Age, Continuous: Age | years | 15.7 (SD 4.1) | 16.2 (SD 4.5) | — | — | 16.0 (SD 4.3) |
| Sex: Female, Male: Female | Participants | 17 | 20 | — | — | 37 |
| Sex: Female, Male: Male | Participants | 0 | 0 | — | — | 0 |
| Region of Enrollment: United States | participants | 17 | 20 | — | — | 37 |
6. Primary Endpoint
The registry lists the primary outcome as Peripheral Insulin Sensitivity at Baseline and 6 Months., measured over Baseline and 6 months.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | Standard Error |
| Unit | mg/kg/min per uU/mL |
| Endpoint classification | Continuous Endpoint |
The primary outcome is peripheral insulin sensitivity, assessed at baseline and 6 months and expressed in mg/kg/min per uU/mL. The registry record says it was evaluated during a hyperinsulinemic-euglycemic clamp. This is a controlled physiological procedure: in general, insulin is administered while glucose is adjusted to maintain a target level, and the resulting measurements are used to assess insulin sensitivity. The record does not provide further procedural details.
The endpoint is relevant to the stated study purpose, which includes investigating treatment effects on insulin sensitivity in overweight adolescent females with PCOS. It is a continuous measure, and the registry specifies that results are characterized by a mean with standard error as the dispersion measure. In general, the mean describes the group's average measurement, while the standard error describes the precision of the estimated mean. It does not describe how widely individual participants' measurements vary.
The outcome is listed at baseline and 6 months, but the registry record does not report any values, standard errors, or treatment-group contrast. It also does not state whether the primary comparison was based on follow-up values, baseline-to-follow-up change, or a model using both measurements. Without those results and analysis details, the endpoint definition alone does not establish the direction or magnitude of any treatment-group difference.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record identifies a clinically focused randomized comparison and specifies peripheral insulin sensitivity as the primary outcome. It does not include the primary outcome means at baseline or 6 months, standard errors, a treatment contrast, a confidence interval, or a p-value. Consequently, the posted record does not allow a reader to assess the estimated treatment difference or its precision. It does not support a conclusion about whether either treatment changed insulin sensitivity or whether the treatments differed.
The record lists 23 outcome measures and names secondary measures that include total fat mass, percent body fat, total testosterone, free testosterone, SHBG, DHEAS, delta androstenedione and delta DHEA at baseline and 6 months. These outcome names indicate the range of measurements registered, not the results. In the absence of outcome values and analysis results, they cannot be used to infer a treatment effect on body composition, androgen measures, or any other domain.
The study's stated purpose also mentions inflammatory markers, vascular markers and bone development. The primary outcome description and secondary outcome list do not provide results for those domains. Their results, measurement definitions and analyses are not reported in the registry record. Similarly, the record does not give participant-level distributions or information that would show whether means represent the measurements across all randomized participants or a subset.
Interpretation of a treatment comparison depends on more than whether a result is labelled statistically significant. In general, the estimated difference describes the observed contrast, while a confidence interval indicates the range of values compatible with the data and the model's assumptions. A p-value addresses compatibility with a specified null hypothesis under an analysis model; it does not measure the size or clinical importance of an effect. None of these inferential results is reported for the primary outcome in this registry record.
Even if a numerical difference were available, the exact estimand would matter. A comparison of values at 6 months could answer a different statistical question from a comparison of changes from baseline, particularly if baseline values differed. The record gives two assessment times but does not report the primary analysis method or exact contrast. It therefore cannot show how baseline measurements were incorporated or whether any adjustment was made.
The safety figures are also limited to the registry's recorded event counts and at-risk denominators. They do not provide detailed event narratives, severity beyond the serious-event category, exposure duration, or a comparative statistical analysis. The absence of reported affected participants in the listed safety fields is not evidence that an event could not occur, and it cannot substitute for the missing efficacy results.
In short, the registry record describes what the study set out to assess, but the outcome and analysis information needed to quantify treatment differences is not reported in the registry record. The design labels and endpoint definitions should not be mistaken for evidence of benefit, lack of benefit, or equivalence between treatments.
9. Statistical Methods Explained
The registry identifies a randomized, parallel-group study with quadruple masking. Randomization is a design method intended to reduce systematic differences between treatment groups at assignment. In general, it supports causal comparisons when assignment is implemented as planned and outcomes are adequately observed and analyzed. The registry record does not describe the sequence-generation procedure, allocation concealment, or the roles included in its masking designation, so those implementation details cannot be evaluated from the record.
What does a mean with standard error describe?
The primary outcome is continuous and its summary parameter is the mean. A mean gives the average value for a group under the chosen analysis. Standard error describes the precision of an estimated mean, with smaller uncertainty generally corresponding to a more precise estimate. It is not a measure of the spread of individual values; a standard deviation or distributional display would address individual variability more directly. The registry lists standard error but reports no primary outcome means or standard errors.
For treatment comparisons, a difference in means would express the contrast on the outcome's measurement scale. A confidence interval would communicate uncertainty around that contrast. General interpretation requires considering both the estimated effect and its uncertainty, rather than relying only on whether a test crosses a conventional significance threshold. No treatment contrast, confidence interval or p-value for this endpoint is reported in the registry record.
How should the two assessment times be understood?
The endpoint is assessed at baseline and 6 months. Repeated measurement can help describe starting values and subsequent values, but the comparison depends on the analysis chosen. A follow-up comparison, a change-score comparison, or a model that uses both baseline and follow-up data does not necessarily target the same quantity. The record does not specify which approach was used, how baseline values were handled, or how within-participant correlation was modeled.
In general, repeated observations from the same participant are related rather than independent. A suitable analysis accounts for that relationship when estimating uncertainty. The record does not provide enough methodological detail to determine whether or how this was addressed. It also does not report a model specification, covariates, transformations, diagnostics, or an analysis population.
What do the sample-size assumptions mean?
The registry describes a planning illustration based on the observed effect, a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation. It says that the effect to detect is a stated fraction of the observed effect. Such calculations are planning illustrations, not observed trial results and not evidence of observed power. A calculation based on an observed effect depends on that effect estimate and the assumptions used; it does not establish the probability that the completed trial detected a true effect.
The record does not report a trial-specific sample-size justification, the target effect used in the original design, or the basis for its enrollment target. The planning assumptions described in the registry should therefore not be treated as the original protocol's design calculation. No calculated planning figures are reported in the registry record.
What can the analysis record establish?
The registry reports 0 analyses. This describes the analysis entries in the registry record; it does not establish that investigators performed no analyses. The record does not report the statistical analysis plan, model used for the primary endpoint, handling of missing observations, or whether analyses were adjusted for baseline values or other covariates. It also does not describe any interim analyses, stopping rules, or sensitivity analyses.
Over-dispersion is a general concern when data vary more than a statistical model assumes, particularly for count outcomes. The primary endpoint is continuous, and no count model is specified for it. More broadly, choosing a model requires considering the measurement scale and data distribution. The registry does not supply distributions or diagnostics that would allow an assessment of model fit for the continuous outcome.
The record describes a multi-outcome study and names several measures beyond the primary endpoint. General statistical practice considers multiplicity when many outcomes or comparisons are assessed, because repeated testing can increase the chance of apparently notable findings by chance. The registry does not report a multiplicity adjustment or a hierarchy of confirmatory and exploratory tests. Since test results are not reported, no adjustment method or adjusted inference can be evaluated.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists 23 outcome measures and identifies several secondary measures, including body-composition and androgen-related measurements at baseline and 6 months. Assessing multiple outcomes or comparisons raises a general multiplicity concern: as the number of statistical tests grows, the chance of at least one apparently notable result arising by chance can increase. This does not mean that every outcome must be adjusted in the same way; the appropriate approach depends on the prespecified aims and analysis plan. That plan and any multiplicity procedure are not reported in the registry record.
The registry record reports 0 analyses. It does not state which outcomes were analyzed, whether the primary outcome was treated as confirmatory, or how secondary outcomes were interpreted. It also does not report p-values or confidence intervals for the listed measures. Without that information, the scope of testing and the strength of evidence across outcomes cannot be assessed.
Missing-data handling is not reported in the registry record. The record does not state whether all randomized participants had measurements at both assessment times, how incomplete observations were treated, or whether any imputation or sensitivity analysis was used. The analysis population is also not reported: it is not possible to determine whether the results, if any, were intended to include all randomized participants or a defined subset.
The registry lists enrollment as 65 and gives safety denominators for the two treatment groups, but those entries do not provide a participant-flow account or explain missing endpoint observations. The record does not report exclusions after randomization, withdrawals, protocol deviations, or reasons for missing measurements. These omissions limit assessment of whether missingness could have affected the estimated comparison.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Rosiglitazone | 0 | 17 | 0.0% |
| Drospirenone/Ethinyl Estradiol | 0 | 20 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Rosiglitazone | 0 of 17 |
| Drospirenone/Ethinyl Estradiol | 0 of 20 |
| Overweight/Obese Without PCOS | 0 of 0 |
| Lean Without PCOS | 0 of 0 |
For rosiglitazone, the registry records 0 participants affected by serious adverse events among 17 at risk, and 0 affected by other adverse events among 17 at risk. For drospirenone/ethinyl estradiol, it records 0 affected by serious adverse events among 20 at risk, and 0 affected by other adverse events among 20 at risk. The two non-PCOS comparison arms have 0 at risk in these fields, consistent with the statement that no participants were enrolled in those arms.
These figures are counts as recorded, not estimates of comparative risk. The registry lists no serious adverse event terms. Death counts are not reported in the registry record. The record does not give event timing, duration of exposure, severity details for other adverse events, or a statistical comparison between treatments. A zero count in the listed fields means no affected participants were recorded in those entries; it does not demonstrate that the treatments carry no risk or establish that the groups are equally safe.
Adverse-event interpretation also depends on how events were collected and coded, whether follow-up was complete, and the duration of observation. Those details are not reported in the registry record. The safety entries should therefore be read as limited registry counts rather than a comprehensive account of all possible harms.
12. What the Result Does, and Does Not, Mean
The record establishes that the study was designed as a randomized parallel comparison of rosiglitazone and drospirenone/ethinyl estradiol in treatment-naive overweight adolescent females with PCOS. It identifies peripheral insulin sensitivity at baseline and 6 months as the primary outcome and names additional measures related to body composition and androgen levels. These details clarify the intended focus of the research.
The record does not establish the size, direction, or precision of any treatment effect. It contains no primary outcome values, treatment contrast, confidence interval, or p-value. A study's design and registered outcome definitions do not by themselves show that a treatment changed a measure, that one treatment differed from the other, or that there was no difference.
Nor does the absence of posted primary results establish equivalence. Demonstrating equivalence or non-inferiority generally requires a specified margin and an analysis designed to assess that question. Such a margin and analysis are not reported in the registry record. A lack of reported statistical significance, even if it were known, would not alone establish that treatments are equivalent.
The results cannot be generalized from the record to the non-PCOS groups, because no participants were enrolled in those arms. They also cannot be extended to populations or treatment circumstances not described by the study's enrolled treatment groups. The record does not provide details of dose, adherence, or post-treatment follow-up that would support broader interpretation.
The listed adverse-event counts have a narrower meaning: no affected participants are recorded in the specified serious or other event fields for the enrolled treatment arms. This does not prove the absence of harm, and the event record does not provide enough detail to compare safety comprehensively. Overall, the posted information describes the study question and some design and safety fields, but leaves the central efficacy results and analysis methods unreported.
13. Why This Trial Matters Statistically
This study is statistically notable for asking a treatment-comparison question using a continuous physiological outcome measured at baseline and 6 months. Such a design can support an informative comparison when the analysis clearly defines the treatment contrast, accounts for repeated measurements, and reports both effect estimates and uncertainty. The registry record identifies the measurement and timing but does not report the analysis that would connect them to a treatment effect.
The comparison is active treatment versus active treatment, so interpretation would depend on the relative outcome estimates rather than a contrast with placebo or no treatment. Randomization is an important design feature for reducing confounding, while transparent reporting is needed to evaluate the resulting estimate. The record's randomized designation is useful context, but the absence of outcome estimates and analysis details prevents statistical interpretation of the treatment comparison.
The presence of multiple registered outcomes also illustrates why a clear analysis hierarchy matters. When a study investigates insulin sensitivity alongside body-composition, androgen, inflammatory, vascular and bone-related measures, readers need to know which analyses were prespecified and how multiplicity was addressed. The registry record does not report that information or provide results for those outcomes.
The record also distinguishes enrolled treatment groups from planned contextual groups: no participants were enrolled in the two non-PCOS arms. This is important when interpreting the study's scope. Those arms cannot supply data on normal developmental changes or serve as cardiovascular-marker controls in the enrolled sample.
Finally, outcome definitions and registry safety counts are not substitutes for a full results analysis. Reporting group estimates, uncertainty, analysis populations, missing-data handling and event details would be needed to judge the strength and limitations of the evidence. Their absence makes this record useful for understanding the study's intended design, but insufficient for estimating treatment effects.
14. 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 central limitation is that the registry record does not report primary outcome results. Means, standard errors and a treatment comparison for peripheral insulin sensitivity are absent, as are confidence intervals and p-values. The main study question therefore cannot be answered from the posted record.
The outcome's timing is specified as baseline and 6 months, but the analysis contrast is not. The record does not clarify whether the primary comparison used 6-month values, change from baseline, or a repeated-measures model. It also does not describe the statistical model, baseline adjustment, covariates, or assumptions. This prevents evaluation of how the endpoint was analyzed.
Important design and conduct details are not reported in the registry record, including the randomization procedure, which roles were masked, treatment dose and schedule, adherence, withdrawals, protocol deviations, and the participant flow through follow-up. Missing-data methods and the analysis population are not reported. Without these details, it is not possible to assess attrition or determine how the analyzed participants, if any, relate to those randomized.
The study lists additional outcomes, but the registry provides no results for them and no reported plan for handling multiplicity. Although it describes non-PCOS comparison arms, it states that no participants were enrolled in those arms, limiting the scope of possible comparisons. The safety entries report affected-participant counts but do not give event narratives or exposure details; deaths are not reported in the registry record.
The record's planning description is an illustration based on an observed effect and stated assumptions, not documentation of the original design justification. The trial-specific sample-size rationale is not reported in the registry record. Taken together, missing results and methodological detail make it impossible to judge the magnitude, precision, robustness, or clinical importance of the treatment comparison.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Insulin Resistance, Polycystic Ovary Syndrome, and Bone Research Study — NCT00640224.
- Registry record last updated 2017-11-17; status last verified 2017-10.
- Results first posted 2017-11-17.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The registry identifies a completed randomized, quadruple-masked, parallel Phase 4 study of rosiglitazone versus drospirenone/ethinyl estradiol in treatment-naive overweight adolescent females with PCOS. Enrollment is listed as 65. The primary outcome is mean peripheral insulin sensitivity, measured during a hyperinsulinemic-euglycemic clamp at baseline and 6 months, with standard error listed as the dispersion measure.
The record lists 23 outcome measures and describes two non-PCOS comparison arms in which no participants were enrolled. It records 0 serious and 0 other adverse-event-affected participants in each enrolled treatment group's safety fields; deaths are not reported in the registry record.
Primary outcome values, treatment contrasts, inferential results, analysis methods, missing-data handling, and the analysis population are not reported in the registry record. The posted information therefore describes the study's question and selected design and safety details, but does not quantify the treatment effects.