Sitagliptin Dose Comparison Study in Patients With Type 2 Diabetes (MK-0431-077)(COMPLETED): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Sitagliptin Dose Comparison Study in Patients With Type 2 Diabetes (MK-0431-077)(COMPLETED) (NCT00541229), a randomized phase 1 study. The primary outcome is 24-hour Weighted Mean Glucose (WMG).

2. Clinical Question

The endpoint reflects glucose measured across a full 24-hour period, summarized as a weighted mean. The registry reports least-squares means with 95% confidence intervals for the treatment groups. Its posted ANOVA comparisons estimate mean differences in final values between sitagliptin 200 mg and placebo, and between sitagliptin 100 mg and placebo.

The registry also lists a comparison between the two sitagliptin doses. That result is reported with an 80% confidence interval and no p-value. The record does not state enough about that comparison’s method or decision criteria to treat it as a conventional superiority test or to establish a non-inferiority conclusion.

The trial was a crossover study, but efficacy results are restricted to Treatment Period I because the registry reports that a carry-over effect was observed between treatment periods. Accordingly, the primary results should be understood as comparisons based on the first treatment period, not as a summary of results across both treatment periods.

3. Trial Design

This was a Phase 1b, multicenter, randomized, double-blind, placebo-controlled crossover clinical trial. It evaluated sitagliptin 100 mg, sitagliptin 200 mg, and placebo in patients with type 2 diabetes whose glycemic control was inadequate on diet and exercise. The registry reports enrollment of 103 participants. The trial is listed as completed, with a start date of 2007-08-24 and primary completion on 2008-06-17.

A crossover design means that participants receive treatments in separate periods rather than remaining in a single parallel treatment group throughout. The registry record does not report the treatment sequence, period lengths, washout details, or the method used to assess carry-over. It does state that a carry-over effect was observed between treatment periods. For that reason, the efficacy results are presented from Treatment Period I only.

The primary endpoint was assessed on Day 7 of Treatment Period I. Its analysis population was per-protocol: randomized participants who had a measurement at that time and did not have a major protocol violation. Missing data were not imputed. The arm counts for this endpoint were 28 for sitagliptin 200 mg, 26 for sitagliptin 100 mg, and 30 for placebo. These counts are smaller than the total enrollment, and the registry’s definition of the analysis population explains the eligibility conditions for inclusion; reasons for individual exclusions are not reported in the registry record.

The record identifies the study as randomized and double-blind but does not report the randomization procedure, allocation ratio, or the parties covered by masking. Further details of the protocol, including the planned sample-size justification and any interim analyses, are not reported in the registry record. Enrollment therefore describes the trial as a whole, while the endpoint counts describe the participants included in the posted per-protocol efficacy results.

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: Following a 4-week (wk) diet/exercise (and wash-off period for patients on antihyperglycemic agent \[AHA\]), patients with fasting plasma glucose (FPG) 130-250 mg/dL were eligible to enter a 2-wk placebo run-in period prior to randomization.

Treatment Period I (Week 1)

Reasons for not completing this period:

Washout Period I (Week 2-5)

Reasons for not completing this period:

Treatment Period II (Week 6)

Reasons for not completing this period:

Washout Period II (Week 7-10)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The per-protocol (PP) population consisted of all patients randomized who had a measurement at Day 7 in Treatment Period I and did not have any major protocol violations. Missing data were not imputed.

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.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as 24-hour Weighted Mean Glucose (WMG), measured over Day 7 of Treatment Period I. Due to a carry-over effect that was observed between treatment periods, efficacy results are presented from Treatment Period I only..

The primary endpoint was 24-hour weighted mean glucose (WMG), measured in mg/dL on Day 7 of Treatment Period I. The registry defines WMG as the area under the 24-hour glucose curve, AUC(0-24 hr), divided by 24, using the linear trapezoidal method. In general, an area-under-the-curve summary combines measurements across a time interval, while division by the interval length expresses the result as an average over that period.

A full-day measure can describe glucose across more of the day than a single measurement at one time. It condenses a series of readings into one continuous outcome that can be compared between treatment groups. The registry reports least-squares means and 95% confidence intervals for the groups. A least-squares mean is a model-based estimate of a group mean; details of the model beyond the posted ANOVA comparisons are not reported in the registry record.

The outcome is a continuous measure, so its reported values and mean differences retain the unit mg/dL. The endpoint does not by itself establish how participants felt or functioned, nor does the record provide a threshold for judging whether a difference is clinically important. Its stated timeframe is limited to Day 7 in the first treatment period. Because the registry reports carry-over between periods, the efficacy presentation does not combine the two periods or represent a treatment-period average across the crossover.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, and the registry reports least-squares means with 95% confidence intervals. The sponsor’s two dose-versus-placebo comparisons are described as ANOVA analyses of mean differences in final values. The registry does not provide the full model specification, its covariates, or other details needed to reconstruct the analysis completely. Statistical interpretation should therefore remain tied to the methods and estimates actually reported.

What does a confidence interval express?

In general, a confidence interval communicates uncertainty around an estimated quantity. A narrower interval indicates greater precision under the analysis assumptions; a wider interval indicates less precision. The 95% intervals reported for the group means and dose-versus-placebo contrasts summarize uncertainty for those estimates. They do not imply that every participant’s value falls within the interval, and they do not measure the range of individual responses.

For a mean difference, zero represents no difference between the compared group estimates. In general, an interval that excludes zero is consistent with evidence of a difference at the corresponding two-sided significance level, assuming the analysis and interval are appropriate. The two reported dose-versus-placebo 95% intervals are entirely below zero. Their p-values are 0.004 and <0.001, respectively. These inferential summaries do not answer whether the differences are clinically important; no clinical-importance threshold is reported in the registry record.

How should the dose comparison be read?

The registry reports a sitagliptin 200 mg versus sitagliptin 100 mg mean difference of 10.1 and an 80% confidence interval from 0.61 to 9999999. It supplies no p-value or method for this contrast. An 80% interval uses a different confidence level from the 95% intervals reported for the placebo comparisons, so the intervals should not be compared as though they express the same degree of uncertainty. The unusually large upper bound makes the interval difficult to interpret; the reason is not reported in the registry record.

A non-inferiority or equivalence claim generally requires a prespecified margin and an analysis designed to assess that margin. The registry’s posted information does not state a margin or describe the method for this dose comparison. The listed interval and mean difference alone therefore do not establish that the doses are equivalent or that one is non-inferior to the other.

Why does the crossover design matter?

In general, crossover trials use participants’ experience in more than one treatment period, and analyses may need to account for period and sequence effects. A carry-over effect occurs when the effect of a treatment in one period may influence outcomes in a later period. The registry specifically reports carry-over between treatment periods and consequently presents efficacy results for Treatment Period I only. Restricting results to the first period avoids interpreting the posted efficacy estimates as if they represented an unaffected comparison across both periods. The record does not report the assessment procedure or additional details of the crossover analysis.

What do the analysis population and missing-data statement imply?

The per-protocol population consisted of randomized participants who had a Day 7 measurement in Treatment Period I and no major protocol violations. Missing data were not imputed. In general, a per-protocol analysis can exclude participants who do not meet specified criteria, so its results describe the analyzed group rather than necessarily all randomized participants. The endpoint counts are 28, 26, and 30 across the three treatment groups, while total enrollment was 103. Reasons for non-inclusion and the handling of any other missing or unusable measurements are not reported in the registry record.

Not imputing missing values means that the registry did not replace missing endpoint measurements with modeled or assumed values for this analysis. That approach avoids relying on an imputation model, but the consequences depend on how much data were missing and why. The number and pattern of missing endpoint measurements are not reported in the registry record, so their possible effect on the estimates cannot be assessed from the posted information.

What can planning calculations tell readers?

In general, sample-size calculations estimate how many participants per arm might be needed for a specified power under chosen assumptions. The planning illustrations associated with these results use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the assumed effect is a stated fraction of the observed effect. Such illustrations are not the trial’s own design justification and are not observed power. The registry record does not report the trial’s sample-size justification, and no planning figure should be interpreted as the number actually enrolled or analyzed.

Multiplicity concerns arise when several hypotheses are tested, because the chance of at least one false-positive finding can increase as the number of tests grows. The registry reports two dose-versus-placebo p-values and a separate dose-versus-dose comparison, but an adjustment strategy or multiplicity plan is not reported in the registry record. The posted p-values should therefore be read as reported for their individual comparisons; the record does not establish whether or how family-wise error was controlled.

10. Analyses Reported by the Sponsor

The registry’s sponsor-reported analyses use ANOVA for the two comparisons with placebo. For sitagliptin 200 mg versus placebo, the reported mean difference in final values is -31.8, with a 95% confidence interval from -53.1 to -10.5 and p = 0.004. For sitagliptin 100 mg versus placebo, it is -41.9, with a 95% confidence interval from -63.6 to -20.2 and p < 0.001. These are the sponsor’s inferential results and should be distinguished from simple, unadjusted recalculations based on posted summary statistics.

The registry also reports the group least-squares means and confidence intervals. Recalculations from group summaries, where displayed, are not a replacement for the sponsor’s ANOVA, which may incorporate model details not captured by a simple comparison of the summaries. No discrepancy between sponsor and recalculated results is identified in the registry record. The full sponsor model specification is not reported in the registry record, so the source of any differences in approach cannot be assessed in detail.

The sponsor’s dose-versus-dose result is reported with an 80% confidence interval and no p-value. Its upper bound is 9999999, and its analysis method is not stated. That comparison should not be treated as if it had the same inferential reporting as the two placebo contrasts. The registry record does not report enough information to determine the intended decision rule for the dose comparison.

11. Multiplicity, Missing Data and Analysis Populations

The registry reports three analyses: two ANOVA comparisons of a sitagliptin dose with placebo and one comparison between the two doses. It does not report whether these tests were part of a prespecified hierarchy, whether an adjustment for multiple comparisons was planned, or how multiplicity was handled. Because the adjustment plan is not reported in the registry record, the p-values should be understood as the values posted for the individual analyses rather than as evidence that a particular family-wise error procedure was applied.

The primary endpoint population was per-protocol. It included randomized participants with a Day 7 measurement in Treatment Period I who had no major protocol violations. The registry reports counts of 28 for sitagliptin 200 mg, 26 for sitagliptin 100 mg, and 30 for placebo. Overall enrollment was 103. The record does not provide participant flow or explain the reasons why enrolled participants were not included in the endpoint analysis.

The registry explicitly states that missing data were not imputed. In general, an analysis without imputation uses observed endpoint data and does not fill in missing outcomes. Its interpretability can depend on the amount and causes of missingness. The number of missing measurements and any assessment of whether missingness differed across groups are not reported in the registry record.

The reported efficacy results are limited to Treatment Period I because a carry-over effect was observed between treatment periods. The record does not describe an interim-analysis plan, a detailed statistical analysis plan, or a sample-size justification. The absence of those details means readers cannot assess whether the posted comparisons follow a particular prespecified multiplicity or missing-data strategy beyond the information stated in the registry record.

12. 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 5% reporting threshold.

Most frequent non-serious adverse event terms

The adverse-event figures are reported by treatment exposure. Serious adverse events were recorded for 0 of 78 participants at risk during sitagliptin 200 mg, 0 of 72 at risk during sitagliptin 100 mg, and 0 of 74 at risk during placebo. The registry lists no serious adverse-event terms. These are counts in the posted safety record; they do not establish that an event could not occur or provide a precise estimate of its frequency beyond the observed participants and exposure periods.

Other adverse events were recorded for 4 of 78 participants at risk during sitagliptin 200 mg, 2 of 72 during sitagliptin 100 mg, and 4 of 74 during placebo. These figures are separate from the serious-event counts and should not be combined with them. The registry record does not provide event descriptions or additional detail about severity, timing, or attribution for these other events.

Deaths are not reported in the registry record. Safety counts should be read alongside the crossover design and the treatment-specific numbers at risk, rather than treated as a direct comparison of long-term risks. The record does not report exposure duration for the safety summaries, so the figures do not describe event rates over a common time scale.

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

The posted results show lower estimated 24-hour WMG in each sitagliptin group than in the placebo group at Day 7 of Treatment Period I. The sponsor’s ANOVA comparisons report mean differences below zero, with 95% confidence intervals that do not include zero and p-values of 0.004 and <0.001. The group means and confidence intervals also describe the uncertainty around the estimated endpoint values in the per-protocol population.

These findings do not by themselves establish the clinical importance of the observed differences. The registry record does not report a threshold for a clinically meaningful change in 24-hour WMG. Nor do the reported results establish that the doses are equivalent or that one dose is non-inferior to the other: the dose comparison has an 80% confidence interval from 0.61 to 9999999, no p-value, and no reported method or decision margin.

The results apply to the reported endpoint, time point, analysis population, and first treatment period. They do not represent all 103 enrolled participants, because the primary analysis was per-protocol and included fewer participants in each arm. They also do not summarize efficacy across both periods, since the registry reports a carry-over effect and presents efficacy results only for Treatment Period I.

The record does not report longer-term outcomes, other clinical outcomes, or a protocol-defined approach to interpreting multiple comparisons. The statistical evidence in the posted contrasts should therefore not be extended beyond those comparisons. The safety figures are limited to the adverse events recorded in the registry and the participants at risk for each treatment; they do not resolve uncertainties about longer exposure or uncommon events.

In general, statistical significance and clinical importance answer different questions. The reported p-values and confidence intervals address evidence and precision for the analyzed mean differences under the sponsor’s analysis. They do not, without additional clinical context, establish the size of benefit that matters to patients or support a treatment recommendation.

14. Why This Trial Matters Statistically

This trial is statistically notable because it combines a crossover design with a documented carry-over effect. Crossover designs can allow within-participant comparisons, but carry-over can complicate interpretation when earlier treatment may influence later-period results. The registry responds to that concern by presenting efficacy results only from Treatment Period I. That restriction is important when reading the reported means and comparisons: they are not estimates pooled across the crossover periods.

The record also illustrates why a result’s analysis population matters. Total enrollment is 103, while the primary endpoint is reported for a per-protocol population with arm counts of 28, 26, and 30. The estimates therefore describe participants meeting the stated analysis criteria and having the required measurement, not every participant enrolled. The registry’s statement that missing data were not imputed adds context, while the reasons and amount of missingness remain unreported.

Another statistical point is that different comparisons are reported with different inferential detail. The two dose-versus-placebo analyses include ANOVA, 95% confidence intervals, and p-values. The dose-versus-dose result instead has an 80% confidence interval, no p-value, and no stated method. This makes it especially important not to treat every posted contrast as having the same analysis or evidential status.

Finally, the record reports several comparisons but does not describe multiplicity handling or a clinical-importance threshold. Those omissions do not erase the reported estimates; they limit the claims that can be made from them. Statistical interpretation depends not only on the point estimates, but also on population, design, uncertainty, endpoint timing, and the scope of the analysis.

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 efficacy interpretation is limited by the crossover carry-over issue. The registry states that carry-over between treatment periods was observed and therefore reports efficacy only from Treatment Period I. The record does not explain how carry-over was assessed, what its magnitude was, or whether additional analyses were considered. The posted results should not be generalized to an overall effect across both periods.

The primary endpoint analysis was per-protocol and used only participants with a Day 7 measurement in Treatment Period I and no major protocol violations. The record reports the endpoint counts but does not report participant-level exclusions, reasons for missing measurements, or how these differed by treatment. Missing data were not imputed; the number and pattern of missing outcomes are not reported in the registry record.

The dose-versus-dose comparison is difficult to interpret because the registry provides an 80% confidence interval with an upper bound of 9999999, no p-value, and no analysis method. A non-inferiority margin or equivalence criterion is not reported in the registry record. Accordingly, this contrast cannot be used to establish equivalence or non-inferiority based on the posted details alone.

The registry does not report a multiplicity plan, a sample-size justification, or the complete statistical model specification. It also does not report a threshold for clinical importance. These omissions constrain assessment of the planned inferential framework, the clinical relevance of the mean differences, and the robustness of conclusions to alternative assumptions. The safety record lacks detailed descriptions of other adverse events and does not report deaths. The findings are therefore specific to the endpoints, analysis population, comparisons, and safety information recorded in the registry.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Sitagliptin Dose Comparison Study in Patients With Type 2 Diabetes (MK-0431-077)(COMPLETED) — NCT00541229.
• Registry record last updated 2017-05-12; status last verified 2017-04.
• Results first posted 2009-07-28.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

This completed Phase 1b trial was randomized, double-blind, placebo-controlled, and crossover. It enrolled 103 patients with type 2 diabetes whose glycemic control was inadequate on diet and exercise. The primary endpoint was 24-hour weighted mean glucose on Day 7 of Treatment Period I.

The registry reports per-protocol least-squares means of 159.0 mg/dL for sitagliptin 100 mg, 169.1 mg/dL for sitagliptin 200 mg, and 200.9 mg/dL for placebo. Sponsor-reported ANOVA comparisons with placebo have mean differences of -41.9 and -31.8, respectively, with 95% confidence intervals and p-values reported for both. A dose-versus-dose comparison is also listed, but its method and decision criteria are not reported.

Because the registry reports carry-over between periods, efficacy results are restricted to Treatment Period I. Missing data were not imputed. The safety record reports no serious adverse events in the listed treatment exposures and gives counts for other adverse events; deaths are not reported in the registry record.