Pilot Trial of Naltrexone for Obesity in Women With Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pilot Trial of Naltrexone for Obesity in Women With Schizophrenia (NCT00793780), a randomized na study. The primary outcome is Change in Body Weight From Baseline.

2. Clinical Question

The study asks whether mean body-weight change over 8 weeks differed between women with schizophrenia or schizoaffective disorder assigned to naltrexone 25mg and women assigned to placebo. The registry identifies obesity as a condition of interest and describes the purpose as evaluating the effect of low dose oral naltrexone on body weight.

This distinction matters: the arm-specific changes describe what was observed within each group, while the clinical question concerns the contrast between assigned interventions. A direct comparison and its uncertainty would address that question more directly than a visual comparison of separate arm summaries. Such an analysis is not reported in the registry record, so the posted primary results do not establish the size or precision of a between-group effect.

3. Trial Design

This was a randomized, quadruple-masked, parallel-group pilot trial sponsored by Yale University. The registry lists an enrollment of 24 and records the study as completed. The study dates are 2008-12 for start and 2010-11 for primary completion. The phase is listed as NA. These record details describe the registered study; they do not by themselves establish how many participants contributed to each particular analysis.

The interventions were naltrexone 25mg and placebo. The primary outcome was change in body weight from baseline at 8 weeks. Weight was measured with shoes off to the nearest 0.1 kg. The primary outcome is continuous, and the posted arm summaries are means accompanied by 95% confidence intervals. The record lists secondary outcomes concerning fasting serum glucose, PANSS, insulin, LDL cholesterol, and craving for sweet or rich foods. Numerical results are posted for PANSS and the craving questionnaire; results for the other listed secondary outcomes are not reported in the registry record.

The arm-level primary results are based on 10 participants in the naltrexone group and 11 in the placebo group. The mean change was -3.4 kg with naltrexone, with a 95% confidence interval from -5.16 to -1.65, and 1.37 kg with placebo, with a 95% confidence interval from -0.3 to 3.03. The registry record does not report a direct primary comparison, its confidence interval, or its p-value. It also does not report the analysis population, missing-data handling, or the statistical model used for a primary comparison.

The registry record does not report a sample-size justification, a detailed statistical analysis plan, or interim analyses. As a pilot trial, the posted results are best read as a limited account of the observed outcomes, rather than as a definitive estimate of the treatment contrast.

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.

Population: Overweight women between the ages of 18 and 70 years who met the DSM-IV criteria for schizophrenia or schizoaffective disorder, on the basis of Structured Clinical Interview for DSM-IV. An additional participant dropped out of the study after baseline data was collected.

6. Primary Endpoint

The registry lists the primary outcome as Change in Body Weight From Baseline, measured over 8 weeks.

The primary endpoint is change in body weight from baseline, expressed in kg, at 8 weeks. It measures the difference between a participant’s weight at the starting assessment and the later assessment, rather than weight at a single time point. The registry specifies that weight was measured with shoes off to the nearest 0.1 kg.

Change from baseline is commonly used when the research question concerns whether an outcome shifts during follow-up. It can describe the direction and size of within-group change and can be compared between randomized groups. The between-group contrast is particularly important: both groups may change over time for reasons unrelated to the assigned intervention, and a comparison with placebo helps distinguish shared change from a difference associated with assignment.

The registry reports the mean change and a 95% confidence interval separately for each group. The naltrexone mean change was -3.4 kg, with a 95% confidence interval from -5.16 to -1.65. The placebo mean change was 1.37 kg, with a 95% confidence interval from -0.3 to 3.03. These summaries describe the recorded arm-specific results, but they are not a direct estimate of the difference between arms. The registry record does not report that primary contrast or its uncertainty.

Interpretation also depends on how complete and comparable the measurements were, which participants were included, and how missing follow-up weights were handled. These details are not reported in the registry record. The measurement description clarifies how weight was taken, but it does not provide the broader analysis details needed to judge potential bias or the precision of a direct treatment comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help clarify what the registered results can and cannot establish. The primary outcome is continuous, the available summaries are arm-specific means, and the registry does not report the direct comparison that addresses the primary question. The secondary comparisons are recalculated, unadjusted comparisons based on the posted summaries rather than analyses reported by the registry.

What does a mean change describe?

A mean change summarizes the average difference between baseline and follow-up measurements within a group. Negative and positive values indicate opposite directions on the measurement scale. For the primary outcome, the registry reports mean change in kg at 8 weeks, with means of -3.4 for naltrexone and 1.37 for placebo. The mean is useful because it provides a concise description of the group’s average observed change, but it does not reveal how individual participants’ changes were distributed.

People in a group can have different changes even when the average is stable. A mean alone does not show whether most observations were close to the average or whether the average was influenced by a small number of larger changes. The registry does not report a primary-outcome standard deviation or another distributional summary. The specified measurement procedure—shoes off and rounded to the nearest 0.1 kg—describes the measurement, not the variability of outcomes or the statistical model.

Why is the direct comparison important?

Random assignment makes the contrast between groups the key comparison for the intervention question. A within-group change can reflect many influences, including changes over time that also occur in the placebo group. The between-group estimate compares average change under the two assignments and quantifies how far apart those averages are. The registry reports separate arm means and their confidence intervals, but does not report the primary between-group estimate, its confidence interval, or its p-value.

Separate confidence intervals should not be treated as the confidence interval for a difference. The uncertainty of a contrast depends on the variability and sample sizes in both groups, and on the analysis method. A formal comparison also requires decisions about such matters as the analysis population and any adjustment for baseline measurements. Those primary analysis details are not reported in the registry record. The primary contrast therefore cannot be reconstructed as a registry-reported result from the arm summaries alone.

How should a confidence interval be read?

A confidence interval communicates uncertainty in an estimate under the assumptions of the method used to construct it. A narrower interval generally indicates greater precision; a wider interval indicates that the estimate is less precise. It does not represent a range containing a specified proportion of individual participant outcomes, and it is not a statement that every value inside it is equally plausible.

The primary 95% confidence intervals accompany separate mean changes. They describe uncertainty about those means and do not answer whether the groups differed. For the secondary recalculated comparisons, the intervals are for mean differences. The PANSS comparison is 4.87 units, with a 95% confidence interval from -9.706 to 19.45. The craving-score comparison is -3.1 units, with a 95% confidence interval from -6.62 to 0.42. Both intervals include zero, the value representing no mean difference. This observation does not establish that the groups are equivalent; the intervals also include a range of possible nonzero differences.

What do unadjusted secondary p-values mean?

A p-value is a measure of how unusual data at least as incompatible with a null hypothesis would be, assuming the null hypothesis and the analysis model. It is not the probability that the null hypothesis is true, and it does not measure clinical importance. The secondary unadjusted p-values are 0.513 for PANSS and 0.084 for the craving questionnaire. They are recalculated from posted summary statistics, not identified as registry-reported analyses.

Unadjusted results do not account for the fact that multiple outcomes may be examined. When many outcomes are considered, the chance of obtaining an apparently notable result by chance can increase. The registry record does not report a multiplicity strategy. Thus, individual unadjusted secondary p-values merit cautious interpretation and should not be isolated from the pattern of outcomes, uncertainty, and missing analyses.

What are planning illustrations?

The secondary outcome records include sample-size planning illustrations based on observed effects. These are hypothetical planning quantities, not observed trial results and not evidence that the study had a particular probability of detecting an effect. The stated assumptions are a two-sided significance level of 0.05, equal allocation, and a normal approximation, with planning based on a stated fraction of the observed effect. For example, the PANSS planning illustration gives 194 participants per arm for 80% power when planning for 100% of the observed effect, and 345 participants per arm for 80% power when planning for 75% of the observed effect. The craving questionnaire illustration gives 28 participants per arm for 80% power when planning for 100% of the observed effect, and 50 participants per arm for 80% power when planning for 75% of the observed effect.

These illustrations are sensitive to the estimated effect used as their starting point. In a pilot study, that estimate may be imprecise, and planning around it can produce unstable projections. The registry record does not report that these calculations were used to design the trial. They should therefore be read as illustrative planning scenarios, not as a design justification or a measure of observed power.

What can be said about missing analysis details?

The registry lists a primary endpoint and arm summaries, but does not report the primary comparison or analysis population. It does not report how missing measurements were handled, whether an intention-to-treat approach was used, or whether the analysis adjusted for baseline values. Without these details, readers cannot fully evaluate the primary analysis strategy or its assumptions. The safest interpretation stays close to what is recorded: arm-specific means and confidence intervals for the primary outcome, and recalculated unadjusted comparisons for selected secondary outcomes.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary endpoint and several secondary outcomes, but it does not report how the analysis accounted for examining multiple outcomes. When statistical tests are conducted across multiple endpoints, the opportunity for chance findings can increase. A multiplicity plan may prioritize a primary endpoint, define a testing sequence, adjust thresholds, or otherwise explain how findings should be interpreted. No such strategy is reported in the registry record. The posted secondary p-values are therefore unadjusted and should be treated as exploratory.

The record reports arm-specific results for the primary outcome but no direct primary comparison. It also does not report the primary analysis population, the handling of missing measurements, or whether participants were analyzed according to randomized assignment. These choices matter because they determine which observations contribute to estimates and how incomplete follow-up is addressed. The registry record does not describe imputation, exclusions, or sensitivity analyses, so their use cannot be assumed.

Missing outcome data can affect both the estimate and its precision, especially if the probability of having an observed measurement is related to outcome or group assignment. General statistical methods may address missingness under stated assumptions, but no particular method or assumption is reported for this trial. The arm sizes listed for the primary outcome are 10 for naltrexone and 11 for placebo; the record does not explain whether these reflect all randomized participants or only those with evaluable data.

The registry also does not report a statistical analysis plan, a sample-size justification, or interim analyses. These omissions do not show that such procedures were absent; they mean their details are not reported in the registry record. Without a reported plan, it is difficult to assess whether endpoint definitions, analysis choices, exclusions, or multiplicity procedures were specified before outcome data were examined.

These reporting gaps constrain interpretation rather than prove that the results are biased. They indicate that readers cannot verify the primary comparison, understand the handling of incomplete observations, or assess the inferential status of the secondary analyses from the registry record alone. The most defensible reading is descriptive and cautious.

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 non-serious adverse event terms

The registry lists serious adverse events for both groups. No serious adverse events were recorded among the 10 participants at risk in the naltrexone group, or among the 11 participants at risk in the placebo group. These counts describe the entries in the registry record; they do not establish that serious events could not occur or that the groups have identical safety profiles.

The record also reports other, non-serious adverse events in 9 participants at risk in the naltrexone group and 5 participants at risk in the placebo group. The registry does not provide event terms for the serious events, and no deaths are reported as a numerical result. The death fields are null, so deaths should be described as not reported in the registry record rather than interpreted as a confirmed absence.

Safety counts depend on the definitions, ascertainment period, and completeness of reporting. Those details are not specified in the posted safety summary. The group counts are based on the listed at-risk participants and should not be converted into broader safety conclusions. In a trial of this size, absence of a recorded serious event is not evidence that the risk is zero, and differences in non-serious event counts are descriptive rather than a formal comparative safety analysis.

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

The registry’s primary results describe mean weight change within each randomized group over 8 weeks. The naltrexone arm mean was -3.4 kg, with a 95% confidence interval from -5.16 to -1.65. The placebo arm mean was 1.37 kg, with a 95% confidence interval from -0.3 to 3.03. These values are observations from the posted record and should not be transformed into a new between-group estimate.

The result does not directly quantify how much mean change differed between assigned groups. That comparison, together with its confidence interval and statistical analysis, is not reported in the registry record. Separate arm estimates cannot replace it. The posted summaries therefore do not provide a registry-reported basis for a definitive conclusion about the comparative effect on weight.

The uncertainty around each mean also matters. A confidence interval reflects uncertainty under the method’s assumptions; it does not capture every possible source of bias, missingness, or measurement error. The registry gives a measurement procedure for body weight but does not report the primary analysis population, handling of missing weights, or analysis model. These omissions limit how fully the posted results can be evaluated.

Likewise, the secondary comparisons are recalculated and unadjusted. They are not a complete analysis of all listed secondary outcomes, and the registry does not report a multiplicity strategy. Their results are compatible with uncertainty and should not be treated as confirmatory. The sample-size planning illustrations are hypothetical calculations based on observed effects, not evidence about the trial’s observed power or a record of the design rationale.

The trial is informative as a pilot record because it describes the randomized comparison, the outcome definition, and observed arm summaries. It also identifies where reporting is incomplete. It cannot establish that the intervention is beneficial, ineffective, equivalent to placebo, or safe based on the recorded estimates alone. Such claims would require a reported comparative analysis and fuller information about the conduct and analysis of the study.

13. Why This Trial Matters Statistically

This trial illustrates why the estimand—the quantity a study is trying to estimate—should be kept distinct from the data summaries available in a registry. The primary question concerns the contrast in weight change between randomized groups. The record instead reports a mean and confidence interval for each arm without reporting the direct primary contrast. Both summaries are useful, but they answer a different statistical question from the one implied by the randomized comparison.

It also demonstrates the importance of uncertainty. A group mean can look clear as a point estimate while still having substantial uncertainty, and separate confidence intervals do not constitute an interval for the difference. A reader needs the contrast and its uncertainty to assess the precision of the primary comparison. When that result is absent, calculating a contrast from arm means would create a new analysis rather than recover a registry-reported finding.

The secondary results illustrate another recurring issue: multiple outcomes and unadjusted inference. The record includes several secondary outcomes, but provides numerical results for only some of them and does not report a multiplicity approach. Individual secondary p-values can attract attention even when the analysis plan and broader set of results are unclear. Their interpretation should reflect that context.

Finally, planning calculations based on observed effects are not observed power and do not document the trial’s design rationale. Their values depend on the effect used for planning, while the registry does not report a sample-size justification. This distinction helps prevent retrospective planning illustrations from being mistaken for evidence that the study was adequately powered for its primary question.

For biostatisticians and clinical researchers, the central lesson is practical: outcome definitions, group summaries, direct contrasts, uncertainty, missing-data methods, and multiplicity plans each answer different questions. Transparent reporting of all these elements is needed for a complete interpretation. The available record supports cautious description, while leaving important inferential details unreported.

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 primary reporting limitation is the absence of a direct comparison between naltrexone and placebo for change in body weight. The arm-specific means and 95% confidence intervals do not provide the primary contrast or its uncertainty. The registry also does not report a primary p-value or the statistical model used.

The number of participants contributing to the primary endpoint is limited, with 10 listed in the naltrexone arm and 11 in the placebo arm. The record does not clarify whether these are all randomized participants or participants with observed outcome data. It does not report missing weights, withdrawals, adherence, or how incomplete follow-up was handled. These omissions make it difficult to assess potential attrition effects.

Further limitations include the absence of primary-outcome variability beyond the confidence intervals, no reported analysis population, and no stated baseline-adjustment approach. The record does not provide detailed eligibility criteria or information about participant characteristics beyond the target population. These details would help assess the applicability and robustness of the findings.

The secondary outcome record is incomplete: numerical results are available for PANSS and the craving questionnaire, while results for fasting serum glucose, insulin levels, and LDL cholesterol are not reported in the registry record. The secondary comparisons are unadjusted, and a multiplicity plan is not reported. Consequently, they should not be treated as confirmatory evidence.

The safety summary reports no serious adverse events in either group, but provides limited detail about event definitions, ascertainment, and deaths. The record does not report a sample-size justification, detailed statistical analysis plan, or interim analyses. These are limits of what can be learned from the registry record and should not be taken to prove that the trial lacked those procedures.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Pilot Trial of Naltrexone for Obesity in Women With Schizophrenia — NCT00793780.
• Registry record last updated 2016-10-06; status last verified 2016-08.
• Results first posted 2016-10-06.
• 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

NCT00793780 is a completed randomized, quadruple-masked, parallel pilot trial sponsored by Yale University. It compared naltrexone 25mg with placebo in women with schizophrenia or schizoaffective disorder, with obesity among the listed conditions. Enrollment was 24. The study started in 2008-12, and primary completion was recorded as 2010-11.

The primary endpoint was change in body weight from baseline at 8 weeks. The registry reports arm-specific means and 95% confidence intervals, but not a direct comparison between groups. Secondary outcomes include fasting serum glucose, PANSS, insulin, LDL cholesterol, and craving for sweet or rich foods; numerical summaries are posted for PANSS and the craving questionnaire. The registry record does not report a multiplicity strategy, missing-data approach, analysis population, or primary statistical analysis. The available results are therefore descriptive and incomplete for the main comparative question.