Comparison of Serum Contraceptive Hormone Levels Between Normal Weight and Obese Users of the NuvaRing®: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparison of Serum Contraceptive Hormone Levels Between Normal Weight and Obese Users of the NuvaRing® (NCT00710606), a non-randomized na study. The primary outcome is Mean Serum Concentrations of Etonogestrel and Ethinyl Estradiol.

2. Clinical Question

3. Trial Design

This was a prospective, parallel-group clinical trial with nonrandomized allocation and no masking. The record lists Columbia University as sponsor and classifies the phase as NA. The study began in 2008-06, had primary completion in 2008-12, and is marked completed.

The groups were defined by BMI rather than assigned at random: obese women had BMI 30-39.9, and normal-weight women had BMI 19-24.9. Both groups used NuvaRing. The arm descriptions state that participants received two rings and, during the second cycle of ring use, returned for serial serum hormone measurements and transvaginal ultrasound twice weekly during four weeks of continuous use. The primary endpoint description specifies measurements at Week 3 and Week 6.

The record says 40 women were enrolled. It also states that two withdrew before the study cycle and one woman was excluded because of removal of the contraceptive vaginal ring during the study cycle. The primary endpoint description says 37 women completed follow-up for serum measurements. These statements describe enrollment and follow-up, but the registry record does not give a participant-flow account by group or explain the handling of every measurement.

Because the comparison was nonrandomized, differences between the groups cannot be attributed to BMI alone from the design description. The record does not report how participants were recruited beyond describing them as interested in initiating the ring, nor does it report the rationale for the chosen group definitions or whether other factors were controlled in analysis.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: 17 participants in each group was planned a priori to have 80% power to identify a one standard deviation difference in the mean serum levels of the contraceptive hormones. We enrolled 40 women, 2 withdrew prior to the study cycle. We excluded one woman due to removal of the CVR during the study cycle leaving 18 normal weight and 19 obese women.

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 Mean Serum Concentrations of Etonogestrel and Ethinyl Estradiol, measured over Measurements at Week 3 and Week 6 continuous ring use.

The primary endpoint was the mean serum concentration of etonogestrel and ethinyl estradiol, reported in ng/L. The endpoint description specifies geometric means with 95% confidence intervals and measurements at Week 3 and Week 6 of continuous ring use. These are the two hormones named in the record as being released by the contraceptive ring.

Serum concentration is a pharmacokinetic measure: it describes the amount of a drug or hormone measured in blood at a specified time. In general, such measurements can help characterize exposure during use. The study summary says the investigation focused on whether the ring maintained effective serum steroid concentrations to suppress ovarian activity in obese women. Concentration measurements are not themselves a direct measure of pregnancy prevention, and the primary endpoint is not a pregnancy outcome.

The registry record does not report the geometric mean estimates, their confidence intervals, or a numerical comparison between BMI groups. It therefore does not show how far apart the group estimates were or how precisely those estimates were measured. The record also does not specify whether each hormone and time point was treated as a separate primary comparison or how repeated measurements within a participant entered the analysis.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help frame what this study was designed to measure, while the registry record leaves important details of the actual analysis unspecified. The explanations below describe general statistical concepts and do not imply that a particular unreported method was used.

What does a geometric mean summarize?

In general, a geometric mean summarizes measurements on a multiplicative scale. It is calculated from the average of logarithmically transformed values and then returned to the original scale. This approach can be suitable when values are positive and the scientific question concerns proportional changes or relative differences. Unlike an arithmetic mean, it is less influenced by very large values when a distribution is strongly right-skewed. Whether that description fits the observed hormone concentrations cannot be checked from the registry record, because the distributions and individual measurements are not reported.

The primary endpoint names geometric means for both hormones and specifies a 95% confidence interval. The record does not explain whether the interval was calculated on a transformed scale, what statistical model was used, or whether the interval describes each group separately or a between-group contrast. These distinctions matter: intervals around separate group estimates are not the same as an interval directly estimating the difference between groups.

What does a confidence interval communicate?

In general, a confidence interval conveys sampling uncertainty for an estimate under its specified statistical procedure and assumptions. A narrower interval indicates greater precision under that procedure; a wider interval indicates less precision. The interval does not display the range of participant-level values, and it does not on its own establish clinical importance. For a between-group question, the most direct quantity is typically a contrast with its own interval, because that estimates the difference of interest and its uncertainty. The registry record gives no numerical confidence intervals for this endpoint, so readers cannot judge precision or whether plausible contrasts are small or large.

Why does the allocation scheme matter?

In a randomized study, random allocation is intended to make treatment groups comparable on average with respect to both measured and unmeasured factors. Here, allocation was nonrandomized and the groups were defined by BMI category. In general, differences in measured concentrations could reflect BMI-related differences, other participant characteristics, or a combination. Statistical adjustment can address some measured differences when suitable data and a prespecified model are available, but it cannot automatically eliminate unmeasured confounding. The registry record does not report covariate adjustment or a detailed analysis model.

The parallel-group description means that the study compares separate groups. Participants also had serial measurements during ring use, according to the arm descriptions. Repeated observations from the same participant are generally correlated rather than independent. In general, an analysis of repeated measurements should account for that correlation or use a prespecified summary for each participant. The registry record does not explain how repeated observations were handled, how many measurements were available per participant, or whether the Week 3 and Week 6 measurements were analyzed separately or jointly.

What does the planning statement mean?

The record states that 17 participants in each group was planned a priori to have 80% power to identify a one standard deviation difference in mean serum hormone levels. In general, a power calculation uses assumptions about the target difference, variability, significance threshold, allocation, and analysis. It describes the chance of detecting a specified effect under those assumptions if that effect is true. It does not show that the study had a particular probability of detecting its actual effect, and it is not a measure of the strength or validity of an observed result.

The planning target should also be distinguished from the number enrolled and the number with follow-up measurements. The registry states that 40 women were enrolled and describes two withdrawals before the study cycle and one exclusion after ring removal; it states that 37 completed follow-up for serum measurements. It does not report a revised power calculation, the variability used in planning, or whether the planning target was met in the analyzed population. The separate sample-size material is described as an illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such illustrations are planning aids, not observed power. Their numerical outputs and the effect estimates needed to evaluate them are not reported in the registry record.

What cannot be inferred without the analysis details?

The registry lists zero analyses and does not report a statistical analysis plan. It does not state which model generated the geometric means, whether the analysis was adjusted, how multiple hormone and time-point comparisons were handled, or how missing measurements were treated. It also does not identify an analysis population beyond saying that 37 women completed follow-up for serum concentrations. In general, these choices affect estimates, uncertainty, and the interpretation of statistical tests. Without them and without numerical results, readers cannot reproduce the analysis or assess whether the reported method matches the study question.

Over-dispersion is a general issue in count data: it occurs when variability exceeds what a simple model assumes and can make standard errors too small if ignored. It is not clear that over-dispersion is relevant to this primary continuous concentration endpoint, and the registry record does not report a count-data model for it. This illustrates why method descriptions should be tied to the data type and analysis actually used rather than inferred from the endpoint label.

10. Multiplicity, Missing Data and Analysis Populations

The primary endpoint includes two hormones and measurements at Week 3 and Week 6. The registry record does not say whether these were treated as separate primary tests, combined into a single analysis, or arranged in a hierarchy. It does not report a multiplicity adjustment or explain how the family of comparisons was defined. In general, testing multiple hypotheses can increase the chance of at least one false-positive finding when each test is interpreted separately. Adjustment or a prespecified hierarchy can address that issue, but the record does not establish that either was used.

Two secondary outcomes are listed: the number of participants achieving a maximum follicle diameter > 13mm during three weeks of follow-up and mean endometrial proliferation. The registry record does not report their results or how these secondary outcomes were related to the primary endpoint in the analysis plan. The number of outcomes alone is not enough to determine the multiplicity strategy.

Missing-data details are limited. The record states that 40 women were enrolled, two withdrew before the study cycle, and one woman was excluded because of ring removal during the study cycle. It also states that 37 completed follow-up for serum concentration measurements. The registry record does not describe the reasons for withdrawal, the availability of individual measurements at each time point, or any imputation or other method for handling missing data.

The analysis population is not fully specified in the registry record. It does not say whether all enrolled participants were analyzed, whether analyses were restricted to completers, or how the exclusion after ring removal was applied to each endpoint. Nor does it report a participant-flow breakdown by group. These omissions limit assessment of selection effects and reproducibility.

11. Safety Results

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

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: fractured pelvis (5.0%), venous thrombosis (5.0%).

The safety record lists 20 participants at risk in each group. It reports zero participants with serious adverse events in the obese group and two in the normal-weight group. The registry record does not give a causal assessment, timing, severity detail beyond the serious-event classification, or a description of how these events related to ring use. These counts alone cannot establish that the ring caused an event or that one group had a higher underlying risk.

The listed serious adverse event terms are fractured pelvis and venous thrombosis, each shown with 5.0%. The record does not make clear in the details given whether those terms are linked to a particular group, whether they represent participants or event reports, or whether a participant could contribute more than one term. They should therefore be read as listed safety terms, not as evidence of a treatment-related pattern.

In general, safety counts from a small study are descriptive and can be imprecise, especially for uncommon events. The registry record does not provide comparative safety analyses or confidence intervals. The reported figures do not establish a difference in risk between the BMI groups, and they should not be interpreted as a causal comparison.

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

This record describes a prospective, nonrandomized comparison of serum concentrations of etonogestrel and ethinyl estradiol among obese and normal-weight NuvaRing users. It establishes the study question, BMI group definitions, the planned endpoint summaries, and the schedule described for continuous ring use. It also records enrollment and follow-up information, along with selected serious adverse event counts.

It does not report the numerical geometric mean concentrations, 95% confidence intervals, or a statistical comparison for either hormone. As a result, the registry record cannot show the direction, size, or precision of a primary endpoint difference. It does not establish whether concentrations were similar or different between the groups, and it does not report a conclusion about the primary question.

The endpoint is a pharmacokinetic measure, not a direct pregnancy outcome. The study summary gives a rationale focused on serum steroid concentrations and ovarian activity, but the primary endpoint described is hormone concentration. The record does not report pregnancy rates or a direct estimate of contraceptive effectiveness for the participants in this study. Concentration findings, even if available, would need to be interpreted as concentration findings rather than treated as direct evidence about pregnancy outcomes.

The nonrandomized design also limits causal interpretation. BMI category defined the comparison groups; participants were not randomized to BMI status. The record does not report adjustment for potential confounding, so a group difference could not automatically be attributed to BMI alone. Conversely, the absence of numerical results does not show that the groups had identical concentrations.

The record supports a description of what was intended and how the groups were observed, but not a conclusion about whether the ring works or does not work, nor about comparative effectiveness. The registry record also does not report numerical results for the secondary outcomes. Any broader clinical interpretation would require results and analytical details that are not reported in the registry record.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of how a clinical question can be framed around a measurable biological endpoint while leaving key interpretive details dependent on reporting. It compares geometric mean serum concentrations across BMI-defined groups and specifies a confidence-interval approach. Those choices signal that the investigators were interested in hormone exposure, but the absence of numerical estimates means readers cannot evaluate the magnitude or precision of the comparison.

The study also illustrates the distinction between a randomized comparison and an observational contrast within a prospective trial. Prospective measurement can establish a planned schedule and collect data over time, but it does not make allocation random. Because the groups were defined by BMI category, group differences may reflect factors beyond BMI; the record does not report whether those factors were measured or addressed analytically.

The repeated-measurement schedule makes the analysis description important. In general, measurements taken from the same participant over time are related, so an analysis must account for that dependence or define a participant-level summary. The registry identifies Week 3 and Week 6 as time points but does not explain how serial measurements were incorporated. Reporting that detail would help readers understand what the endpoint estimate represents.

The stated planning target is also useful context, but it should not be confused with evidence about the observed result. The record gives a target of 17 participants in each group for 80% power to identify a one standard deviation difference. That statement does not show that the analysis achieved its intended precision, especially when the numerical outcome estimates and their uncertainty are not reported.

Finally, the gap between a biological endpoint and a clinical outcome matters. Serum concentrations address exposure, while the record does not report pregnancy outcomes for this study. Keeping those outcomes distinct prevents a pharmacokinetic comparison from being interpreted as a direct estimate of contraceptive effectiveness.

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.

A central limitation of the registry record is that it does not report the primary endpoint results. The stated endpoint, summary measure, confidence-interval level, and time points are available, but the group estimates and their uncertainty are not. Without these results, the main comparison cannot be quantified or independently interpreted.

The nonrandomized design limits causal conclusions. Participants were grouped by BMI category, and the registry record does not describe adjustment for potential confounders or provide enough detail to assess baseline comparability. The record also does not report how participants were recruited beyond saying they were interested in initiating the ring.

Follow-up information is brief. The record reports enrollment of 40 women, two withdrawals before the study cycle, one exclusion after ring removal, and 37 women completing serum measurement follow-up. It does not provide a detailed flow by group, explain the reasons for withdrawals, or specify how missing observations were handled. The analysis population and treatment of repeated measurements are not reported.

Statistical reporting is limited: no analyses are listed, and the registry record does not describe the model, covariate adjustment, multiplicity strategy, or handling of missing data. The planned power statement is available, but the variability assumptions and detailed calculation are not. No numerical primary or secondary outcome estimates are reported, limiting assessment of both effect size and precision.

Finally, the primary endpoint is serum concentration, not pregnancy or another direct measure of contraceptive effectiveness. The registry record does not report pregnancy outcomes for this trial, so its endpoint cannot directly answer that clinical outcome question.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Comparison of Serum Contraceptive Hormone Levels Between Normal Weight and Obese Users of the NuvaRing® — NCT00710606.
• Registry record last updated 2015-04-30; status last verified 2015-04.
• Results first posted 2015-04-30.
• 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.

18. Record Summary

NCT00710606 is a completed, prospective, nonrandomized, parallel-group study sponsored by Columbia University. It compared obese women with BMI 30-39.9 and normal-weight women with BMI 19-24.9, with both groups using NuvaRing. The record lists no masking and classifies the phase as NA.

The primary endpoint was geometric mean serum concentrations of etonogestrel and ethinyl estradiol in ng/L, with 95% confidence intervals at Week 3 and Week 6 of continuous ring use. The record states that 40 women were enrolled and that 37 completed follow-up for serum measurements. It describes two withdrawals before the study cycle and one exclusion after ring removal.

The registry record does not report numerical primary or secondary outcome results, or a statistical analysis plan. Safety listings show zero serious adverse events among 20 obese participants at risk and two among 20 normal-weight participants at risk. These descriptive counts do not establish causation or a comparative risk.