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 Bone Loss in Women With Anorexia Nervosa (NCT00089843), a randomized phase 2, phase 3 study. The primary outcome is Bone Mineral Density.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00089843 |
| Phase | Phase 2, Phase 3 |
| Condition | Anorexia Nervosa |
| Allocation | Randomized |
| Design model | Factorial |
| Masking | Quadruple |
| Enrollment | 77 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Massachusetts General Hospital |
| Results posted | Yes |
2. Clinical Question
Population
Women with anorexia nervosa. The registry record does not report further eligibility details in the posted summary, such as age criteria or how participants’ bone status was defined.
Intervention
Risedronate, listed as Actonel, and testosterone were studied in a factorial design. Participants were assigned to combinations of active risedronate or its placebo and active testosterone or its placebo.
Comparator
The comparison for each intervention is its placebo condition within the factorial design. The four assigned combinations were placebo for both interventions, active testosterone with placebo risedronate, active risedronate with placebo testosterone, and both active interventions.
Primary question
How did percent change in postero-anterior spine bone mineral density over 12 months vary with risedronate and with testosterone, when each intervention’s effect was assessed separately regardless of whether a participant also received the other intervention? The registry record reports factorial results rather than a direct comparison of each combination against the placebo-only combination.
The study rationale in the registry record is that women with anorexia nervosa have been found to have low bone density. The stated aim was to determine whether low doses of a natural hormone, testosterone, and/or risedronate, a medication described as helping prevent bone breakdown, would improve or prevent bone loss in this condition. These are the study’s motivating questions, not conclusions about the posted results.
The main outcome is a measurement of bone mineral density, not a measure of fracture risk or symptoms. The registry also lists markers of bone metabolism as a secondary outcome. The posted primary results therefore address a specific bone-density measure over a defined follow-up period; they do not by themselves answer every clinical question raised by the study rationale.
3. Trial Design
The registry record describes a completed randomized factorial study, with phase listed as Phase 2 and Phase 3. It lists enrollment as 77 and identifies Massachusetts General Hospital as the sponsor. The recorded start is 2003-06 and primary completion is 2008-04. The masking field says quadruple; the record does not explain which roles were masked or how masking was maintained.
Factorial assignment combined the two study interventions. The four listed combinations were placebo Actonel with active testosterone patch; active Actonel with active testosterone patch; active Actonel with placebo testosterone; and placebo testosterone patch with placebo Actonel. This design lets investigators assess the intervention effects separately while participants may receive either combination of active and placebo assignments. The registry says its factorial analysis determines the effect of each intervention separately, whether or not a subject received the second intervention.
The primary endpoint was assessed from baseline to 12 months. The analysis population description says one subject was excluded and that data from all 76 subjects who participated were used to determine the effect of each intervention. It does not state the reason for exclusion or provide further detail about the relationship between that description and the group counts attached to the posted estimates.
Details such as the randomization sequence, allocation concealment, eligibility criteria beyond the condition, treatment adherence, and handling of participants who stopped treatment are not reported in the registry record. The record also does not describe the intended sample-size calculation or the operational details of follow-up. These gaps limit what can be reconstructed about trial conduct from the registry entry 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
| Stage | Placebo Actonel and Active Testosterone Patch | Active Actonel and Active Testosterone Patch | Active Actonel and Placebo Testosterone | Placebo Testosterone Patch and Placebo Actonel |
|---|---|---|---|---|
| STARTED | 19 | 20 | 20 | 18 |
| COMPLETED | 17 | 16 | 15 | 11 |
| NOT COMPLETED | 2 | 4 | 5 | 7 |
Reasons for not completing this period:
| Reason | Placebo Actonel and Active Testosterone Patch | Active Actonel and Active Testosterone Patch | Active Actonel and Placebo Testosterone | Placebo Testosterone Patch and Placebo Actonel |
|---|---|---|---|---|
| Withdrawal by Subject | 1 | 4 | 5 | 6 |
| Physician Decision | 1 | 0 | 0 | 1 |
Analysis population for the primary outcome, as defined in the registry: 1 subject was excluded from analysis. A factorial analysis was performed and determines the effect of each intervention separately, whether or not a subject received the 2nd intervention. Therefore, data from all 76 subjects who participated were used to determine the effect of each intervention on our endpoints.
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 | Placebo Actonel and Active Testosterone Patch | Active Actonel and Active Testosterone Patch | Active Actonel and Placebo Testosterone | Placebo Testosterone Patch and Placebo Actonel | Total |
|---|---|---|---|---|---|---|
| Participants analysed | 19 | 20 | 20 | 18 | 77 | |
| Age, Categorical: <=18 years | Participants | 0 | 0 | 0 | 0 | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 19 | 20 | 20 | 18 | 77 |
| Age, Categorical: >=65 years | Participants | 0 | 0 | 0 | 0 | 0 |
| Age, Continuous | years | 27.1 (SD 7.3) | 25.2 (SD 6.2) | 25.3 (SD 6.3) | 26.9 (SD 7.2) | 26.1 (SD 6.7) |
| Sex: Female, Male: Female | Participants | 19 | 20 | 20 | 18 | 77 |
| Sex: Female, Male: Male | Participants | 0 | 0 | 0 | 0 | 0 |
| Region of Enrollment: United States | participants | 19 | 20 | 20 | 18 | 77 |
6. Primary Endpoint
The registry lists the primary outcome as Bone Mineral Density, measured over Baseline and 12 months.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | 95% Confidence Interval |
| Unit | percent change |
| Endpoint classification | Continuous Endpoint |
The primary endpoint is percent change in postero-anterior (PA) spine bone mineral density, measured by dual energy x-ray absorptiometry (DXA), from baseline over 12 months. Bone mineral density is a continuous measurement of bone mass in the scanned region. In this trial, the chosen region was the PA spine.
The registry describes the analysis as using differences in log-transformed values, reported as percent change. Log transformation is a way of expressing a change on a relative scale rather than simply comparing raw measurement units. The posted endpoint is therefore a percent-change summary, not an absolute change in DXA units. The registry lists the mean as the summary parameter and a 95% confidence interval as its dispersion measure.
As a general statistical point, bone mineral density can serve as a measurable biological outcome when a study is investigating changes in bone over time. A continuous measure can capture a range of changes rather than only classifying participants into categories. However, a change in a measured density value is not itself a direct count of fractures, symptoms, or other clinical events. The registry record does not report a fracture endpoint as the primary outcome.
The posted results concern the stated spine measure and follow-up interval. The registry record does not report additional details about DXA acquisition, quality control, or how repeated measurements were handled beyond the description of log-transformed values and percent change.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry reports a factorial estimate for risedronate of 3.2% change in PA spine bone mineral density, with a 95% confidence interval from 1.8% to 4.6% and p < 0.001. It reports a factorial estimate for testosterone of -0.6%, with a 95% confidence interval from -2.0% to 0.8% and p = 0.41. Both results are for percent change between 0 and 12 months, and both are identified as factorial analyses.
These are separate intervention-specific findings, not a direct head-to-head comparison of the two drugs and not results for the four combinations as independent treatment arms. The registry explains that each intervention’s effect was assessed separately, whether or not a participant received the other intervention. The estimates should therefore be read in the context of the factorial analysis, rather than as a comparison between the risedronate value and the testosterone value.
For risedronate, the reported confidence interval is entirely above zero. The small p-value is consistent with evidence against a zero mean percent change under the analysis used. For testosterone, the confidence interval includes zero, and the reported p-value is 0.41. This does not establish that the true change is exactly zero, nor does it show that all potentially relevant effects have been ruled out. The interval is compatible with values on either side of zero within its stated limits.
A confidence interval communicates the range of estimates reasonably compatible with the analysis and its assumptions. It reflects statistical uncertainty around the reported mean, not the range of individual participant outcomes. A p-value measures how unusual the observed data would be under a specified null model; it is not the probability that the null hypothesis is true, and it does not describe clinical importance.
The registry lists group counts of 40 for the risedronate estimate and 39 for the testosterone estimate. These values are attached to the posted summaries, while the population description separately says one subject was excluded and that data from all 76 participants who participated informed the factorial assessment. The record does not explain the apparent difference in how these analysis counts are presented. The group counts should not be treated as a direct description of the four assigned combinations or used to infer why the estimates have those values.
The endpoint is a surrogate measurement of spine bone density over 12 months. The results do not, on their own, establish what happened to fractures, long-term bone health, or outcomes outside the measured region and period. They describe the posted analysis and its statistical precision, not a general conclusion for every person with anorexia nervosa.
The record lists one secondary outcome, markers of bone metabolism, and reports two outcome measures and four analyses overall. Detailed results for those secondary measures are not given in the registry record for this interpretation. Their absence means the primary findings should not be read as a complete account of every measured outcome.
9. Statistical Methods Explained
The key design feature is the factorial analysis. The registry says that the effects of risedronate and testosterone were estimated separately, regardless of whether participants also received the other intervention. This is different from treating the four treatment combinations only as four unrelated groups. The analysis aims to estimate an intervention-specific result while incorporating the factorial assignment structure.
What does a factorial estimate represent?
In general, a factorial design studies more than one intervention in the same trial. When the effects are assessed separately, the estimate for one intervention can reflect information across the levels of the other intervention. This can make efficient use of a study’s assignments, but interpreting the estimate depends on the design, the model, and whether the intervention effects interact. The registry record does not report whether an interaction was assessed, what model specification was used, or how any interaction would affect interpretation.
The results are presented as percent change between 0 and 12 months. The registry says differences in log-transformed values are reported as percent change. General statistical interpretation: a transformation can make an analysis more appropriate for the distribution or scale of a measurement, while conversion to percent change can make results easier to interpret on a relative scale. The record does not give the exact formula used to convert the fitted results to the reported percentages.
How should the confidence intervals be read?
The reported 95% confidence intervals are 1.8% to 4.6% for risedronate and -2.0% to 0.8% for testosterone. In general, a confidence interval describes uncertainty in an estimated population quantity under the model and sampling assumptions. It is not a range that contains 95% of individual participants’ measurements. Nor does it account automatically for every possible source of uncertainty, such as measurement problems or departures from the analysis assumptions.
The interval for the risedronate estimate lies above zero, while the interval for testosterone includes zero. This describes the placement of the posted intervals relative to a zero percent-change reference. It should not be converted into a statement about whether an individual participant’s bone density increased or decreased: the estimates are means, and individual responses can vary. The registry does not report a dispersion measure for individual outcomes in the posted arm summaries.
What do the p-values add?
The p-values are reported as p < 0.001 for risedronate and p = 0.41 for testosterone. In general, a p-value summarizes the incompatibility of the observed data with a specified null hypothesis, given the model. It does not quantify the size or clinical importance of an effect. A larger p-value is not proof of no effect; it can arise when the data are compatible with a range of possible values or when the estimate is imprecise. The confidence interval is useful alongside the p-value because it displays the reported range of uncertainty.
The registry labels both analyses as factorial and gives the parameter as percent change between 0 and 12 months. It does not report whether the p-values are adjusted for multiple outcomes or tests, whether the analysis was prespecified, or whether any covariates were included. Details of the statistical model, standard-error estimation, missing-data assumptions, and treatment of repeated measurements are also not reported in the registry record.
Why does over-dispersion matter in general?
Over-dispersion occurs when observed variation is greater than a statistical model expects. It is especially relevant for some count outcomes, where a simple model may understate uncertainty if the data are more variable than assumed. The primary outcome is continuous, and the registry does not report an over-dispersion assessment or a method intended to address it. It would be inappropriate to infer that over-dispersion was present or that a particular adjustment was used. More generally, any model must represent the variability and structure of the data well enough for its uncertainty estimates to be credible.
The registry’s sample-size assumptions describe a planning illustration based on an observed effect, with two-sided significance level 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect expressed as a fraction of the observed effect. This is a planning exercise, not observed power and not evidence of the trial’s original design rationale. The registry record does not report the trial’s own sample-size justification. For time-to-event outcomes, the described illustration would use a total event count; for other outcomes it would use participants per arm. The primary endpoint is continuous, but the specific planning values are not included in the registry record for this explanation.
The record lists two outcome measures and four analyses, but it does not provide a full statistical analysis plan or explain how analyses were selected for reporting. General statistical practice distinguishes a prespecified primary analysis from exploratory analyses and considers multiplicity when several outcomes or hypotheses are examined. Without the registry’s analysis plan, the intended hierarchy, multiplicity handling, and role of any secondary analyses are not reported in the registry record.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Factorial analysis | Percent change between 0 and 12 months 3.2 | 1.8 to 4.6 | < 0.001 |
| Factorial analysis | Percent change between 0 and 12 months -0.6 | -2.0 to 0.8 | 0.410 |
The posted primary results are the registry’s reported factorial analyses: 3.2% for risedronate, with a 95% confidence interval from 1.8% to 4.6% and p < 0.001, and -0.6% for testosterone, with a 95% confidence interval from -2.0% to 0.8% and p = 0.41. The estimates are stated as percent change between 0 and 12 months. They are presented as intervention-specific factorial results, not as direct comparisons between the four assigned combinations.
No separate recalculated direct two-arm comparison is available for this continuous endpoint in the registry record. The listed analysis is factorial, and the record identifies two-arm comparison values as missing or not estimable. The displayed estimates should therefore not be reconstructed by comparing the posted intervention-specific means or by treating those means as a treatment-versus-placebo contrast.
The discrepancy flag is false, so there is no reported discrepancy between a sponsor analysis and a simple recalculation to resolve. The registry record does not report a separate recalculated estimate to set against the sponsor’s analysis. Interpretation should remain anchored to the sponsor’s posted factorial results and their stated confidence intervals and p-values; those results do not supply a head-to-head effect between risedronate and testosterone.
11. Multiplicity, Missing Data and Analysis Populations
The record identifies Bone Mineral Density as the primary outcome and lists Markers of Bone Metabolism as a secondary outcome. It also reports two outcome measures and four analyses. The registry does not explain how the analyses correspond to the outcome measures, whether all planned analyses are reported, or whether a hierarchy was used to distinguish confirmatory from exploratory testing.
Multiplicity arises when several outcomes, treatment effects, time points, or statistical tests are considered. In general, examining multiple hypotheses can increase the chance of at least one apparently noteworthy result even when none of the tested effects is real. Researchers may address this through prespecified hierarchies, adjusted significance thresholds, or other procedures. Which approach, if any, was used for this trial is not reported in the registry record. The posted p-values should be read as reported, without assuming an adjustment or assuming that none was needed.
The population description states that one subject was excluded from analysis and that data from all 76 subjects who participated were used to determine the effect of each intervention. The registry does not report why the subject was excluded, whether the exclusion occurred before or after treatment assignment, or how missing outcome measurements were handled. It does not say whether the analysis followed an intention-to-treat principle, whether participants were analyzed according to assignment, or whether a per-protocol population was also considered.
The group counts attached to the outcome summaries are 40 for risedronate and 39 for testosterone. The record does not explain how these counts relate to the population statement or whether they represent the same analysis set for each intervention. They should not be used to infer the allocation pattern or the handling of missing measurements. The registry also does not report imputation, sensitivity analyses for missingness, or assumptions about outcomes not observed at the 12-month assessment.
These omissions matter because missingness and choices among analyses can affect both the estimate and its uncertainty. Without details of the analysis plan and missing-data methods, the extent to which the posted results reflect the originally intended analyses cannot be determined from the registry record.
12. 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 |
|---|---|---|---|
| Placebo Actonel and Active Testosterone Patch | 6 | 19 | 31.6% |
| Active Actonel and Active Testosterone Patch | 1 | 20 | 5.0% |
| Active Actonel and Placebo Testosterone | 5 | 20 | 25.0% |
| Placebo Testosterone Patch and Placebo Actonel | 2 | 18 | 11.1% |
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 |
|---|---|
| Placebo Actonel and Active Testosterone Patch | 12 of 19 |
| Active Actonel and Active Testosterone Patch | 16 of 20 |
| Active Actonel and Placebo Testosterone | 16 of 20 |
| Placebo Testosterone Patch and Placebo Actonel | 12 of 18 |
Most frequent serious adverse event terms
| Event term | Organ system | Placebo Actonel and Active Testosterone Patch | Active Actonel and Active Testosterone Patch | Active Actonel and Placebo Testosterone | Placebo Testosterone Patch and Placebo Actonel |
|---|---|---|---|---|---|
| Hospitalization related to underlying condition (anorexia nervosa) | Psychiatric disorders | 6 of 19 | 1 of 20 | 5 of 20 | 2 of 18 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Placebo Actonel and Active Testosterone Patch | Active Actonel and Active Testosterone Patch | Active Actonel and Placebo Testosterone | Placebo Testosterone Patch and Placebo Actonel |
|---|---|---|---|---|---|
| Mild increase in acne or oily skin | Skin and subcutaneous tissue disorders | 8 of 19 | 13 of 20 | 13 of 20 | 12 of 18 |
| Mild irritation at patch site | Skin and subcutaneous tissue disorders | 9 of 19 | 11 of 20 | 8 of 20 | 7 of 18 |
| Mild increase in body hair growth | Skin and subcutaneous tissue disorders | 6 of 19 | 7 of 20 | 3 of 20 | 3 of 18 |
| Reflux | Gastrointestinal disorders | 5 of 19 | 3 of 20 | 1 of 20 | 6 of 18 |
| Myalgias | Musculoskeletal and connective tissue disorders | 3 of 19 | 1 of 20 | 2 of 20 | 0 of 18 |
Serious events with the highest arm-level rate: hospitalization related to underlying condition (anorexia nervosa) (31.6%).
The registry lists serious adverse events by assigned combination. The reported counts affected are 6 among 19 at risk for placebo Actonel with active testosterone patch; 1 among 20 at risk for active Actonel with active testosterone patch; 5 among 20 at risk for active Actonel with placebo testosterone; and 2 among 18 at risk for placebo testosterone patch with placebo Actonel. These are group-specific counts, not a factorial estimate of an intervention’s effect on serious adverse events.
The most common serious-event term listed is hospitalization related to the underlying condition, anorexia nervosa, reported as 31.6%. The registry record does not give further detail about the definition, timing, or causal attribution of this event term. A serious adverse event is not necessarily an event caused by a study intervention; the record’s listing alone does not establish causation.
Other adverse events are also listed: 12 among 19 at risk, 16 among 20 at risk, 16 among 20 at risk, and 12 among 18 at risk for the same respective combinations. These counts should be interpreted separately from the serious-event counts. The record does not define the full set of events included in the other-adverse-event category in the posted summary.
Death counts are not reported in the registry record. The counts and percentages above are descriptive safety reporting and do not, on their own, establish comparative risk. The record does not provide a statistical comparison of adverse events or enough detail to attribute events to either intervention.
13. What the Result Does, and Does Not, Mean
The central statistical result is that the posted factorial estimate for risedronate is a mean 3.2% change in PA spine bone mineral density over 12 months, with a 95% confidence interval from 1.8% to 4.6% and p < 0.001. The estimate for testosterone is -0.6%, with a 95% confidence interval from -2.0% to 0.8% and p = 0.41. The confidence intervals describe uncertainty around the estimates under the analysis used; they do not describe individual responses.
The risedronate interval is above zero, while the testosterone interval includes zero. These are statistical descriptions of the posted summaries. They do not prove that the risedronate result is clinically important, that testosterone has no effect, or that the two interventions differ from one another. The record reports separate factorial analyses, not a direct head-to-head test of the two intervention effects.
The endpoint is a surrogate measure of bone density at the PA spine and is assessed over 12 months. The posted results do not establish what the interventions do to fractures, symptoms, or outcomes beyond the measured site and follow-up. They also do not establish how the results would apply to people outside the population described in the registry. Details on age, disease severity, eligibility, and other characteristics are not reported in the registry record.
The factorial design permits assessment of each intervention separately, but important interpretive details are missing. The record does not report whether an interaction between risedronate and testosterone was tested, how multiplicity was handled, or how missing observations were addressed. The population description notes one exclusion, but does not give its reason. Those omissions make it difficult to assess the robustness of the estimates from the registry information alone.
Safety counts are reported across the four combinations, including serious-event counts and other adverse-event counts. They are descriptive and do not identify which intervention, if any, caused an event. Death counts are not reported in the registry record. The safety information should therefore be considered alongside the limits of the outcome reporting rather than as a definitive comparison of treatment risks.
Overall, the registry gives a specific estimate for a measured bone-density outcome and a separate estimate for each study intervention. It does not supply evidence for every broader clinical question raised by the trial rationale, nor does it provide enough methodological detail to independently reconstruct the analysis.
14. Why This Trial Matters Statistically
This study is statistically notable for using a factorial design to examine risedronate and testosterone within the same trial. A factorial structure can address separate intervention questions while including assigned combinations of active and placebo treatments. The registry explicitly describes its analysis as estimating the effect of each intervention separately, regardless of whether a participant received the other intervention.
That structure also makes clear why the reported results must be read carefully. The values are not four independent estimates for the four combinations, and they are not a direct comparison of risedronate against testosterone. They represent intervention-specific factorial results. The registry does not say whether interaction between the interventions was examined, so the record does not support a fuller assessment of whether the effect of one intervention depended on the other.
The endpoint illustrates the difference between a measurable biological outcome and a patient-important event. PA spine bone mineral density can be measured continuously using DXA, and the registry reports percent change over 12 months. Such a result can describe change in a specified anatomical measure, but it cannot be substituted for an unreported fracture outcome or assumed to represent long-term effects.
The reported confidence intervals and p-values also show why estimation and hypothesis testing are best read together. The intervals show the precision of the estimated mean change within the model, while p-values address compatibility with a null hypothesis. Neither alone establishes clinical importance, and neither describes the spread of participant-level responses.
Finally, the registry’s limited methodological detail highlights the importance of reporting the analysis population, missing-data methods, interaction assessments, and multiplicity strategy. Without those details, readers can describe the posted estimates accurately but cannot fully judge the analysis choices that produced them.
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 description leaves several trial-specific uncertainties. Although it lists randomized allocation and quadruple masking, it does not describe the randomization procedure, allocation concealment, which people were masked, or whether masking was maintained. It also does not give detailed eligibility criteria or participant characteristics beyond the condition being studied and the description of women with anorexia nervosa.
The analysis population description says one subject was excluded and that data from all 76 subjects who participated were used for the factorial assessment. The reason for exclusion, timing of exclusion, and relation to the assigned combinations are not reported. The counts attached to the two outcome summaries differ, and the registry does not explain the relationship between those values and the population description. Missing outcome data, treatment discontinuation, adherence, and sensitivity analyses are not described.
The statistical reporting is also incomplete for independent evaluation. The record names a factorial analysis and reports mean percent change with 95% confidence intervals and p-values, but it does not provide the full model, covariate adjustments, standard-error method, or any assessment of interaction between the interventions. Whether the analysis was prespecified and how multiple outcomes and analyses were handled are not reported in the registry record.
The primary measure covers PA spine bone mineral density from baseline to 12 months. Further details of measurement procedures and quality assurance are not reported. The registry lists markers of bone metabolism as a secondary outcome, but the details and results of those measures are not included in the registry record used for this account. The posted endpoint does not establish effects on fracture outcomes or longer-term bone health.
For safety, adverse-event counts are reported by combination, but the registry does not give enough information to infer causation or compare intervention-specific risks. Death counts are not reported in the registry record. These limits mean that the posted results can be described, but they cannot resolve all questions about trial conduct, analysis robustness, or clinical consequences.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Bone Loss in Women With Anorexia Nervosa — NCT00089843.
- Registry record last updated 2020-09-02; status last verified 2020-08.
- Results first posted 2012-06-01.
- 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.
19. Record Summary
The completed study, NCT00089843, is titled Bone Loss in Women With Anorexia Nervosa. The official title is IGF-1 and Bone Loss in Women Anorexia Nervosa. The registry lists Phase 2 and Phase 3, randomized allocation, quadruple masking, and a factorial model. Massachusetts General Hospital is listed as sponsor, enrollment is 77, the start is 2003-06, and primary completion is 2008-04.
The primary endpoint is mean percent change in PA spine bone mineral density measured by DXA from baseline to 12 months. The factorial analysis reports 3.2% for risedronate, with a 95% confidence interval from 1.8% to 4.6% and p < 0.001, and -0.6% for testosterone, with a 95% confidence interval from -2.0% to 0.8% and p = 0.41. The registry also lists markers of bone metabolism as a secondary outcome.
The record describes a subject exclusion but does not report its reason or the missing-data approach. It does not report a full statistical analysis plan, multiplicity handling, or an intervention-interaction assessment. Safety information includes serious and other adverse-event counts by assigned combination; death counts are not reported in the registry record.