Synthetic vs Natural Estrogen in Combined Oral Contraception: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Synthetic vs Natural Estrogen in Combined Oral Contraception (NCT02352090), a randomized phase 4 study. The primary outcome is Matsuda Index- Whole Body Insulin Sensitivity Index.

2. Clinical Question

The study's broader objective was to compare metabolic effects of natural estradiol and synthetic ethinylestradiol in combined oral contraception, using a progestin-only preparation for comparison. The stated interests also included glucose metabolism, coagulation and markers of chronic inflammation. The hypothesis described in the record was that natural estradiol would influence glucose, coagulation and inflammation markers less than ethinylestradiol, and that the progestin-only preparation would not affect those parameters. These are hypotheses, not findings established by the primary outcome.

Fasting insulin and several lipid, inflammation and coagulation measures appear among the secondary outcomes. The primary result should therefore be read as evidence about the reported Matsuda-index endpoint, not as a complete answer to every metabolic or biological question posed by the study.

3. Trial Design

This was a randomized, phase 4, parallel-group study with no masking reported. The registry record lists Helsinki University Central Hospital as the sponsor and records the study as completed. The recorded start date is 2015-04-01, and the primary completion date is 2018-12-30. The registry enrollment is 59.

Three intervention arms are described: synthetic estrogen plus progestin (ethinyl estradiol / dienogest), natural estrogen plus progestin (estradiol valerate / dienogest), and progestin-only (dienogest). The primary outcome summaries report group counts of 19, 20 and 17, respectively. These are the counts attached to the reported outcome estimates; the registry record does not explain how those outcome counts relate to the enrollment figure or specify the analysis population used.

Randomization indicates that assignment was intended to be allocated by chance, but details such as sequence generation, concealment, stratification or block procedures are not reported in the registry record. The study was unmasked, so participants and study personnel were not reported as blinded to the assigned preparation. The record does not describe the practical consequences of this design for outcome assessment.

The primary endpoint was the change in Matsuda index from baseline to 9 weeks. Arm-level means and 95% confidence intervals are reported, along with an overall ANOVA p-value. The registry record does not report the ANOVA model details, adjustment factors, planned pairwise comparisons, a detailed statistical analysis plan or the handling of missing observations. The number of outcome measures is recorded as 11, and nine secondary outcomes are named. Further details on their definitions and analysis are not reported in the registry record.

The study's short stated primary-outcome timeframe means these results address change over the recorded period, not longer-term patterns. The registry record does not report treatment doses, adherence, co-interventions or the schedule of measurements beyond the primary outcome timeframe. Those omissions limit how fully the intervention procedures can be reconstructed from the record alone.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

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 Matsuda Index- Whole Body Insulin Sensitivity Index, measured over We calculated the change in Matsuda index from baseline to 9 weeks..

The primary endpoint is the Matsuda Index-Whole Body Insulin Sensitivity Index, reported as a mean change from baseline to 9 weeks. The Matsuda index is calculated from glucose and insulin measurements during a standard 2h Oral Glucose Tolerance Test and corresponding fasting values. It is used as an index of insulin sensitivity, meaning the body's responsiveness to insulin in handling glucose.

The registry description says the Matsuda index is correlated with the rate of whole-body glucose disposal during the euglycemic insulin clamp. It also states that lower values are associated with greater insulin resistance, and that a decrease in the index over the study period would indicate decreased insulin sensitivity. This provides a direction for interpreting change: a negative mean change represents a decline in the index over the measured period, while a positive change represents an increase.

The endpoint is relevant to the study's stated interest in glucose metabolism because it condenses glucose and insulin measurements into a continuous measure. Its interpretation remains specific to this index and timeframe. It does not, by itself, report clinical outcomes or fully represent the study's other interests in coagulation, inflammation or lipids. The registry record does not report a threshold for a clinically important change, so the practical importance of any observed change cannot be determined from the index estimates alone.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports a continuous primary outcome summarized as a mean change in Matsuda index. A mean is the arithmetic average of the observed changes in a group. It gives a concise account of the group's central tendency, but it does not show the distribution of individual responses. Participants may have changes above or below the group mean, and a group mean alone cannot show how variable those changes were.

What do the confidence intervals communicate?

Each arm's mean change is accompanied by a 95% confidence interval. In general, a confidence interval expresses uncertainty in an estimate: a narrower interval indicates greater precision, while a wider interval indicates less precision. It is not a range that contains every participant's outcome, and it is not a guarantee that a particular interval captures the true value. Its interpretation depends on the statistical method and assumptions used to construct it.

For the synthetic-estrogen-plus-progestin arm, the estimate is -1.02 and the interval is -2.51 to 0.46. For natural estrogen plus progestin, the estimate is -0.1 and the interval is -1.34 to 1.14. For progestin-only, the estimate is -1.45 and the interval is -3.22 to 0.32. Because each interval spans zero, the arm-level summaries do not precisely distinguish a negative mean change from no change or a positive mean change. These intervals concern changes within arms; they should not be substituted for intervals around between-arm contrasts.

What does the ANOVA test?

The reported analysis is an ANOVA involving all three arms, with p = 0.27. In general, analysis of variance compares group means by assessing the variation between groups relative to variation within groups. Its null hypothesis is that the group means are equal. The reported p-value describes how compatible the observed test result is with that null model, under the test assumptions; it is not the probability that the null hypothesis is true.

The registry record does not report the ANOVA test statistic, degrees of freedom, model specification, diagnostics or adjustment variables. It also does not say whether pairwise comparisons were planned or how they would have been adjusted for multiple testing. Accordingly, the overall p-value should be interpreted as the reported global test, not as a set of pairwise conclusions. A global test can indicate whether the observed means collectively depart from a common-mean model, but it does not identify which groups differ.

Why do assumptions matter?

In general, ANOVA relies on assumptions about the observations and the error distribution, including independence and an appropriate model for the variability within groups. The registry record does not report diagnostics that would allow readers to assess those assumptions for this analysis. The outcome is continuous, and the record reports group means and confidence intervals, but it does not provide dispersion measures beyond those intervals or individual observations.

Random allocation supports a comparison of assigned groups when the design and analysis are implemented appropriately. It does not remove every potential source of uncertainty. In an unmasked study, knowledge of assignment may affect behavior, co-interventions or measurement processes, depending on the outcome and procedures. The registry record does not describe whether or how such possibilities were addressed. The endpoint is based on laboratory measurements, but no further measurement-quality information is reported.

What is not available for re-analysis?

The record supplies arm summaries and a three-group ANOVA result, but not the individual observations, standard deviations, covariance information or detailed analysis model. The record also identifies values as missing or not estimable for a two-arm comparison of this continuous endpoint. Without the relevant contrast estimates or underlying data, a numerical two-arm re-calculation cannot be reported from the available summary. A visual or verbal comparison of arm means is not a substitute for a formal pairwise test.

The record does not report whether the analysis followed intention-to-treat, per-protocol or another population definition. It does not explain how participants with incomplete measurements were handled, whether imputation was used, or whether the analysis was adjusted for baseline Matsuda index. These methodological details matter because they can affect the estimate and its uncertainty. No model or assumptions should be inferred where the registry record is silent.

How should power illustrations be read?

General sample-size planning uses assumptions about the effect to be detected, variability, significance level, allocation and statistical method. The planning assumptions recorded for this study use a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation, with the target effect expressed as a fraction of the observed effect. Such planning values, when displayed, are illustrations based on those assumptions; they are not observed trial results and do not establish the study's original design rationale. The registry record does not report the trial's own sample-size justification or planned power calculation.

10. Analyses Reported by the Sponsor

The sponsor's posted primary analysis is an ANOVA across the three intervention arms, with p = 0.27. The record also reports mean changes and 95% confidence intervals separately for each arm. Those arm summaries describe change within each group; they do not directly supply a confidence interval or p-value for a particular pairwise difference.

The registry record indicates that values for a two-arm comparison of this continuous endpoint are missing or not estimable. Consequently, there is no two-arm recalculation to set alongside the sponsor's overall analysis. The appropriate interpretation is therefore based on the sponsor's reported ANOVA and the posted arm-level summaries, without constructing an unadjusted contrast from unavailable information. No discrepancy between a sponsor result and a recalculated result is reported.

The global ANOVA addresses whether the three means are consistent with a common mean under the test's assumptions. It does not answer every pairwise question, and the record does not report the additional details needed to reconstruct the analysis independently. The sponsor's p-value should not be expanded into claims about unreported pairwise effects or other outcomes.

11. Multiplicity, Missing Data and Analysis Populations

The registry record reports one primary outcome and an overall ANOVA analysis. It names nine secondary outcomes: fasting insulin, high-sensitivity C Reactive Protein, total cholesterol, LDL, HDL, triglyceride, D-dimer and F1+2. The list contains fewer named outcomes than the recorded number of outcome measures, and the record does not clarify the remaining measures or how all outcomes were handled analytically.

When a study tests multiple outcomes or multiple comparisons, the chance of obtaining at least one apparently unusual result can increase. General approaches include identifying a primary outcome in advance, adjusting significance thresholds, or treating secondary analyses as exploratory. The registry record does not report a multiplicity procedure, a hierarchy of testing, or whether pairwise comparisons were planned. It is therefore not possible to determine from the record whether or how the overall false-positive risk was controlled across all outcomes and analyses.

The record does not report an analysis population definition, missing-data counts for the primary endpoint, reasons for missing measurements, or a method for handling missingness. It also does not state whether all randomized participants were included in the primary analysis. The reported arm counts should not automatically be interpreted as the number randomized, the number completing follow-up, or the number analyzed under a particular population definition.

In general, missing outcomes can affect both the precision and the potential bias of an estimate, depending on how much information is missing and why it is missing. No assumptions about the missingness mechanism or imputation approach are warranted. The registry record does not report sensitivity analyses addressing missing data, deviations from assigned treatment or other analysis populations. Those gaps limit assessment of robustness and the extent to which the reported result represents all randomized participants.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The adverse-event figures report zero serious adverse events and zero other adverse events in each of the three groups, with zero deaths in each group. The recorded at-risk counts are 19 for synthetic estrogen plus progestin, 20 for natural estrogen plus progestin and 17 for progestin-only. These are the registry's reported figures, not evidence that no adverse events could occur.

A zero count means that no events of the specified category were recorded in the reported group data. It does not establish that the underlying event risk is zero, particularly when the observed groups are limited in size and the observation period recorded for the primary outcome is 9 weeks. The registry record does not provide further detail on safety follow-up duration, event collection procedures, severity categories beyond serious and other, or whether events were considered related to treatment.

The safety figures should be interpreted separately from the Matsuda index outcome. They do not establish comparative safety between preparations, and they cannot address uncommon or later events from these data alone. The record lists no serious adverse-event terms. This is a description of the posted safety information, not a conclusion about the safety of any preparation.

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

The primary result describes mean change in Matsuda index over the interval from baseline to 9 weeks for three contraceptive preparations. The three arm estimates are negative, and all three 95% confidence intervals include zero. The overall ANOVA p-value is 0.27. Together, these results do not provide strong evidence of a difference among the three group means, but they do not prove that the preparations have equal effects.

The estimate for natural estrogen plus progestin is closer to zero than the estimates for synthetic estrogen plus progestin and progestin-only. That descriptive pattern should not be treated as proof of a difference: the intervals are uncertain, and no pairwise result is reported. A non-significant overall test also does not establish equivalence or non-inferiority. Such conclusions require prespecified margins and an appropriate analysis, neither of which is reported in the registry record.

The endpoint reflects an index of insulin sensitivity, not all dimensions of metabolic health. It does not provide a direct result for coagulation, inflammation or lipid outcomes, even though those topics appear in the study objective and secondary-outcome list. The study's stated hypothesis about natural estradiol and ethinylestradiol therefore cannot be judged from the primary endpoint alone.

Clinical importance is also unresolved. The registry record does not report a threshold for a meaningful Matsuda-index change or a relationship between the observed changes and clinical outcomes. The result should be read as an uncertain short-term comparison of the recorded group means, rather than as a determination of clinical benefit, harm or equivalence. The record does not support treatment recommendations.

14. Why This Trial Matters Statistically

This trial illustrates the distinction between a within-group change and a between-group treatment comparison. The reported arm means describe average change over time within each assigned group. The overall ANOVA addresses whether the group means differ, but it does not identify which specific preparations differ. Keeping those questions separate prevents arm-level movement from being mistaken for evidence of a comparative effect.

The confidence intervals are central to understanding precision. Each arm estimate is accompanied by uncertainty that includes zero, and the intervals cover both negative and positive changes. This makes clear that a point estimate alone can give an incomplete impression. The reported p-value likewise should be read as a result of a specified global test, not as a measure of effect size, clinical importance or probability that the interventions are identical.

The design also raises useful questions about what a registry record can establish. Randomization and a parallel-group structure identify the intended comparison, while the absence of masking and the lack of reported analysis details leave uncertainties about conduct and interpretation. The record does not document a sample-size rationale, multiplicity plan, analysis population or missing-data strategy. These are not reasons to substitute assumptions for evidence; they are reasons to state the limits of the available record.

Finally, a short-term continuous biomarker endpoint can answer a focused statistical question without resolving broader health outcomes. The Matsuda index is relevant to insulin sensitivity, while the study also considered coagulation, inflammation and lipid measures. Clear separation of the primary result from unreported secondary findings keeps conclusions aligned with the evidence actually recorded.

15. 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 gives limited detail about the study procedures and statistical analysis. It does not report eligibility criteria beyond the description of healthy women in the study objective, participant characteristics, treatment doses, adherence, concomitant treatments or measurement schedules beyond the primary timeframe. This makes it difficult to assess how closely the enrolled population and procedures can be characterized from the registry information.

Although the study is recorded as randomized, sequence generation and allocation concealment are not described. Masking is recorded as none, and the record does not explain how outcome measurement or analysis was protected from knowledge of assignment. The primary outcome uses glucose and insulin measurements, but assay methods, reproducibility and quality-control procedures are not reported.

The reported primary outcome includes arm means, confidence intervals and an overall ANOVA p-value, but not the full model specification, test statistic, degrees of freedom or diagnostics. No pairwise contrast estimates are reported. The registry record does not explain the difference between the enrollment figure and the outcome counts attached to the arm summaries, nor does it define the analysis population or describe missing data. These omissions limit independent verification and assessment of robustness.

The record does not report a sample-size justification, a prespecified clinically important difference, the multiplicity strategy or a detailed analysis plan. It also does not provide results for the secondary outcomes in the information used for this analysis. The safety summary is limited to the recorded event counts and does not establish the duration or completeness of safety surveillance. These limitations mean the results should be interpreted narrowly, as the posted short-term primary-outcome summaries and global test.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Synthetic vs Natural Estrogen in Combined Oral Contraception — NCT02352090.
• Registry record last updated 2023-06-22; status last verified 2023-05.
• Results first posted 2021-06-29.
• 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.

19. Record Summary

NCT02352090 is a completed, randomized, unmasked, parallel-group phase 4 study sponsored by Helsinki University Central Hospital. The registry enrollment is 59. It compared ethinyl estradiol / dienogest, estradiol valerate / dienogest and dienogest alone.

The primary outcome was mean change in Matsuda index from baseline to 9 weeks. The arm estimates were -1.02, -0.1 and -1.45, respectively, each with a 95% confidence interval that includes zero. The reported overall ANOVA p-value was 0.27. A two-arm recalculation is not available from the posted summary. The registry record does not report detailed analysis-population, missing-data, multiplicity or sample-size justification information.