Studying the Effects of Sibutramine on Eating Behavior: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Studying the Effects of Sibutramine on Eating Behavior (NCT01170364), a randomized phase 4 study. The primary outcome is 24 Hour Measured Caloric Intake.

2. Clinical Question

This question is about an observed difference in a continuous eating-behavior measure under the two treatment conditions. Because the study used a crossover design, the treatment conditions were experienced by the same participants rather than by separate parallel groups. That makes within-participant comparison relevant to the analysis, although the registry does not report individual paired measurements or the estimated within-participant difference. The posted means describe intake under each condition; they do not by themselves explain how much participants differed from their own measurements across conditions.

3. Trial Design

The registry identifies this as a randomized, double-masked, phase 4 crossover study sponsored by the New York State Psychiatric Institute. It concerns obesity and lists sibutramine and placebo as drug interventions. Enrollment was 14. The registry lists the study start as 2009-02 and primary completion as 2010-10, and its status is terminated. A reason for termination is not reported in the registry record.

In a crossover design, participants receive each treatment condition in a sequence. The descriptions give a sibutramine period of one week and a placebo period of two weeks, with the sequence differing between the treatment descriptions. The primary outcome description compares intake after one week of sibutramine with intake after one week of placebo. The treatment-labeled outcome groups each list 9 participants, and the registry says that all participants received both treatments. These labels therefore represent treatment conditions, not independent parallel groups of participants.

Randomization indicates that treatment assignment or sequence was randomized, and double masking indicates that masking was used; further details of the randomization and masking procedures are not reported in the registry record. The record does not report how treatment order was generated, how allocation was concealed, or who was masked. It also does not describe any washout interval, period effects, or carryover assessment. Those details matter when interpreting crossover results, since measurements under a later condition may be influenced by timing or by an earlier condition.

The record lists one outcome measure and one analysis. Its primary endpoint was 24 Hour Measured Caloric Intake, with a timeframe of one week. The trial is marked terminated, but the registry does not state the reason, whether termination was planned, or whether the posted analysis was affected. A detailed statistical analysis plan and a sample-size justification are not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: 14 of 45 patients were randomized. Of those not randomized, 21 did not meet inclusion criteria.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: This is a cross over design study. All participants received sibutramine and placebo. The sibutramine arm listed here includes all participants who received sibutramine (regardless of whether they received it first or second). The placebo arm included all those who received placebo (regardless of whether they received it first or second).

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 24 Hour Measured Caloric Intake, measured over 1 week.

The primary endpoint was 24-hour measured caloric intake, expressed in kcal. The record describes assessment through 24 hour weighed intake after one week of sibutramine administration compared with 24 hour weighed intake after one week of placebo administration. It is a continuous measure, and the posted treatment summaries use means and standard deviations.

Weighed food intake is used as a direct measure of eating behavior: it records the amount of food consumed over the stated assessment period rather than relying solely on a participant’s recollection or a rating of appetite. The registry summary describes the broader aim as developing assessments of eating behavior, cognitive functioning, and psychological functioning under sibutramine and placebo conditions. The posted primary outcome, however, concerns caloric intake; it is not a measure of all those domains.

The timing is important to the interpretation. The endpoint description specifies intake after one week of each treatment, so it reflects the measurement conditions and period recorded in the trial. It does not establish a longer-term pattern of eating or describe outcomes beyond that assessment. The registry does not report the weighing procedure in further detail, how intake was converted to kcal, or whether any additional dietary measurements were used.

In a crossover study, treatment-specific means can summarize the outcome under each condition, but a paired comparison depends on the within-participant differences. The registry reports a paired test but does not report the average paired difference or its confidence interval. Consequently, the treatment-specific summaries communicate the observed levels and spread, while the inferential result addresses the paired comparison without giving its estimated magnitude or precision directly.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The analysis named in the registry is a paired sample t-test, two tailed. A paired test is suited to data in which each measurement in one condition is linked to a measurement from the same participant in another condition. In this crossover study, the comparison is between caloric-intake measurements under sibutramine and placebo. The analysis therefore focuses on within-participant differences rather than treating the treatment conditions as unrelated groups.

What does pairing contribute?

Participants can differ from one another in their usual intake. Pairing accounts for that person-to-person baseline variation by comparing each participant with their own measurement under the other condition. When measurements within a participant are positively related, the paired analysis can provide a more focused comparison than an analysis that ignores the matching. This is a general property of paired designs, not a claim about the correlation in this trial; that correlation is not reported in the registry record.

The registry reports treatment-specific means and standard deviations, but not the individual pairs or the mean of the paired differences. The treatment means show the average observed intake under each condition, while the standard deviations describe the spread of measurements within each condition. Neither statistic reveals how the two measurements co-vary within participants. As a result, the paired test cannot be reconstructed from the treatment summaries alone. The record also gives no estimated paired difference, standard error, confidence interval, or test statistic.

What does a two-tailed test ask?

A two-tailed test evaluates departures in either direction from the null hypothesis of no average paired difference. It does not restrict attention to a decrease or increase. The registry reports p <0.004 for this test. In general, a p-value measures how unusual the observed test result, or a more extreme result, would be if the null hypothesis and the test assumptions held. It is not a probability that the null hypothesis is correct, and it does not quantify the size or practical importance of a difference.

The t-test relies on assumptions about the distribution of the paired differences. In general, the test is sensitive to the behavior of those differences, rather than to the two treatment distributions considered separately. The registry does not report a distributional assessment, a diagnostic plot, or a robustness analysis. With no individual data or paired-difference summary in the record, readers cannot assess the shape of the differences or independently reproduce the inferential result.

How should the descriptive summaries be read?

The reported means are 1923.0 kcal for sibutramine and 2373.0 kcal for placebo. Their associated standard deviations are 655.0 and 774.0, respectively. A mean gives a measure of the center of observed values, while a standard deviation describes their spread around the mean. Neither one describes uncertainty in the mean in the same way as a confidence interval. The standard deviations are not estimates of the paired treatment contrast, because the within-participant linkage is not encoded in the separate summaries.

A confidence interval for the paired difference would describe the precision of that estimate under the analysis model. No such interval is reported. Its absence means that the registry result does not directly communicate which effect sizes are compatible with the observed paired data. The p-value alone cannot fill that gap: statistical evidence against a null value and precision about an effect estimate are related but distinct features of an analysis.

What does the crossover structure add?

In a crossover design, each participant receives more than one treatment condition, so the treatment comparison can use within-person information. Sequence and period can matter: measurements may differ by when they occur, and an earlier treatment may influence a later measurement. These are general considerations for crossover studies. The registry describes differing treatment sequences but does not report a washout interval, period-specific outcome summaries, a carryover assessment, or an analysis accounting for period or sequence. The reported paired test is identified by name, but the record does not give further model details.

The treatment-specific outcome groups each list 9 participants. Since the design is crossover and the registry says all participants received both treatments, those counts are not evidence of two independent arms. The record does not explain whether every participant had an outcome recorded in both periods or whether the paired test used all enrolled participants. The analysis population and missing-data handling are not reported in the registry record.

How do the planning illustrations differ from trial results?

The planning illustrations are described as based on the observed effect, with equal allocation, a two-sided significance level of 0.05, and a normal approximation. They consider detecting a stated fraction of the observed effect. Such calculations are planning exercises: they explore the number of participants per arm that an assumed design might require under those assumptions. They are not observed power, do not describe what this trial was designed to achieve, and do not replace a report of the original sample-size justification. The trial’s own design assumptions and sample-size justification are not reported in the registry record.

These statistical explanations are general. For this trial, the inferential result available in the registry is the sponsor-reported paired sample t-test, two tailed, with p <0.004. Its interpretation remains conditional on the test assumptions, and the missing paired effect estimate and confidence interval constrain assessment of magnitude and precision.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Sibutramine", "Placebo")
n   <- c(9, 9)
m   <- c(1923, 2373)
sp  <- c(655, 774)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Sibutramine  9  1923.000  655.000  1419.522  2426.478
CARM  Placebo  9  2373.000  774.000  1778.051  2967.949

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as a paired sample t-test, two tailed, with p <0.004. It does not report the test statistic, an estimated paired difference, or a confidence interval. The available treatment means and standard deviations are not enough to recalculate a paired t-test, because the paired analysis requires information about the within-participant differences or their variability. The record therefore does not support an independent recalculation of the reported inferential result from the posted summary statistics.

No recalculated comparison is available to set beside the sponsor’s result. The treatment-specific summaries can be described directly: the mean was 1923.0 kcal under sibutramine and 2373.0 kcal under placebo. Those summaries do not reproduce the paired analysis or establish the paired difference estimate. The sponsor’s reported p-value is the only posted inferential result for the primary comparison, and it should be read as reported rather than as a result independently verified from the summary data.

The registry does not report enough detail to evaluate the analysis beyond its stated method and p-value. In particular, the paired observations, analysis population, and handling of missing outcome measurements are not described. The registry record does not identify a discrepancy between the posted result and a recalculation; rather, the paired result is not recalculable from the summaries available.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and one analysis. It does not report a multiplicity adjustment, a hierarchy of outcomes, or a statistical analysis plan specifying how multiple tests would be handled. Since only one posted analysis is listed, the reported p-value can be described as the result for the listed primary comparison. The record does not establish whether other analyses were planned or conducted but not posted, so the p-value should not be treated as a summary of every possible analysis.

The registry says that all participants received sibutramine and placebo, and the treatment-labeled outcome groups each list 9 participants. It does not report how many participants had usable measurements under both conditions, whether the paired analysis included all enrolled participants, or whether any outcome values were missing. The reason for termination is also not reported. Consequently, the completeness of the primary-outcome data and the population represented in the paired test cannot be assessed from the registry record.

Methods for handling missing observations are not reported in the registry record. There is no description of imputation, exclusion rules, or a sensitivity analysis for missing data. The record also does not report an intention-to-treat or per-protocol population definition. In a crossover study, missing a measurement in either condition can affect whether a participant contributes a complete pair; the registry does not say how this issue was addressed. This is a general methodological concern, not evidence that missing observations occurred in this study.

The study is marked terminated, but the registry does not state why or whether termination affected recruitment, follow-up, or analysis. It would be speculative to attribute the status to the observed result. The record also does not provide interim-analysis details or stopping rules. Those matters are not reported in the registry record and cannot be inferred from the posted p-value or status alone.

13. 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 5% reporting threshold.

Most frequent non-serious adverse event terms

The serious adverse event summary lists 9 participants at risk under sibutramine and 9 under placebo. It reports zero serious adverse events in each treatment condition and zero deaths in each. Because this is a crossover study and the same participants received both treatments, the two treatment summaries should not be interpreted as separate sets of participants.

The registry also lists other adverse events affecting 7 participants under sibutramine and 5 under placebo. These are reported separately from serious adverse events. The record does not provide event terms for serious adverse events, and the listed serious-event terms are empty. It also does not give details about the nature, timing, severity, or duration of the other adverse events.

A zero count in this registry summary means that no serious adverse events were reported in the listed treatment condition; it does not establish that an event could not occur. The summary is limited to the recorded exposure and participants at risk, and the record does not describe how adverse events were collected or whether the categories overlap across treatment periods. These figures are descriptive and should not be treated as a complete characterization of safety.

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

The posted data describe a lower mean 24-hour measured caloric intake under sibutramine than under placebo: 1923.0 kcal compared with 2373.0 kcal. The registry also reports a paired sample t-test, two tailed, with p <0.004. Together, these registry entries describe the direction of the treatment-specific means and the sponsor’s inferential result for the paired comparison. They do not provide the size of the average within-participant difference or its confidence interval.

The p-value does not say how large or important a difference is. It also does not show the probability that the treatment conditions are equivalent or that the result will be repeated in another study. The separate treatment means and standard deviations cannot substitute for a paired effect estimate, because a crossover analysis uses the relationship between each participant’s measurements. The registry does not provide the information needed to estimate that relationship independently.

The outcome is a measurement of 24-hour intake after one week of each condition. The result therefore speaks to this endpoint and timing in this study, not to longer-term eating patterns or other outcomes. Although the study summary mentions cognitive and psychological functioning as aims, the registry lists one primary outcome measure and does not post secondary outcome results. The findings cannot be extended to those domains based on the caloric-intake result alone.

The design provides within-participant exposure to both conditions, but the registry does not describe handling of period effects, carryover, missing measurements, or the analysis population. Those omissions leave uncertainty about important details of how the comparison was implemented and who contributed to it. The study is marked terminated, without a reported reason. No explanation for that status can be inferred from the outcome result.

Thus, the result is a reported statistical comparison accompanied by treatment-specific descriptive summaries. It does not establish the clinical importance of the observed difference, a long-term outcome, or a broader effect beyond the measured endpoint and study conditions. The posted information supports a careful description of what was observed, while leaving the magnitude, precision, and some design and analysis details unresolved.

15. Why This Trial Matters Statistically

This trial is statistically notable because its crossover structure makes the distinction between treatment-specific summaries and paired inference especially important. The means and standard deviations describe each treatment condition, but the paired t-test compares within-participant differences. Without the paired data or their summary, the reported p-value cannot be reconstructed from the separate means and standard deviations. This illustrates why analysis details should be considered alongside descriptive statistics.

The registry reports enrollment of 14, while each treatment-labeled outcome group lists 9 participants and the study description says that all participants received both conditions. These counts refer to treatment conditions in a crossover record, not independent parallel arms. The record does not explain how the enrollment and outcome counts relate or whether every participant contributed a complete pair. Clear reporting of analysis populations and paired-data availability would help readers understand the basis of the comparison.

The result also shows why a p-value and an effect estimate answer different questions. The registry gives p <0.004, but no paired difference or confidence interval. Statistical evidence against a null hypothesis does not by itself communicate the magnitude or precision of the treatment contrast. Readers need both types of information to interpret an estimated effect fully.

Finally, the endpoint is narrowly defined: 24-hour weighed caloric intake after one week of each condition. It is a direct measure relevant to eating behavior, but it does not encompass every outcome mentioned in the study summary. The limited endpoint reporting, together with missing detail on crossover periods, missing data, and analysis populations, makes this record useful as an example of both the value and limits of registry results.

16. 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 main limitation for interpreting the primary comparison is the absence of a reported paired difference estimate and confidence interval. The treatment-specific means and standard deviations show the observed summaries, but they do not reveal the within-participant covariance or the variability of paired differences. This prevents independent reconstruction of the paired t-test from the posted summaries and limits assessment of the estimate’s precision.

The registry says all participants received both treatments, but does not explain why the overall enrollment is 14 while the treatment-specific outcome groups each list 9 participants. It does not state whether the listed counts represent complete outcome data, whether all participants had measurements in both periods, or how participants with missing measurements were handled. The analysis population is not reported.

Further crossover details are absent. The registry does not report a washout interval, period-specific intake summaries, an assessment of carryover, or whether the analysis accounted for sequence or period. These are considerations for evaluating crossover comparisons, but the record does not establish that any particular issue affected this result. The randomization and masking procedures are also not described beyond the labels randomized and double masked.

The registry lists one outcome and one analysis, but does not provide a statistical analysis plan, multiplicity strategy, interim-analysis information, or the trial’s own sample-size justification. The planning illustrations are based on the observed effect and are not evidence of the study’s original design rationale. The reason for terminated status is not reported. Adverse-event summaries give counts but little information about event types or collection methods. These gaps limit evaluation of the result’s precision, completeness, and context.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Studying the Effects of Sibutramine on Eating Behavior — NCT01170364.
• Registry record last updated 2017-08-30; status last verified 2017-06.
• Results first posted 2017-07-31.
• 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.

20. Record Summary

This phase 4 study was randomized, double masked, and used a crossover design. It concerned participants with obesity, was sponsored by the New York State Psychiatric Institute, and had enrollment of 14. The record lists a study start of 2009-02, primary completion of 2010-10, and status of terminated; the reason for termination is not reported.

The primary outcome was 24-hour measured caloric intake after one week of sibutramine compared with one week of placebo. The treatment means were 1923.0 kcal and 2373.0 kcal, respectively, and the standard deviations were 655.0 and 774.0. The registry reports a paired sample t-test, two tailed, with p <0.004. It does not report the paired difference estimate or a confidence interval, and the posted summaries do not allow the paired analysis to be recalculated.

The serious adverse event summary reports zero serious events and zero deaths in each treatment condition. The record lists other adverse events affecting 7 participants under sibutramine and 5 under placebo. Details about missing data, analysis populations, crossover-period effects, and the reason for termination are not reported in the registry record.