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 DPP-4 Inhibition and TZD for DM Prevention (NCT01006018), a randomized na study. The primary outcome is Insulin Secretion.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01006018 |
| Phase | NA |
| Condition | Prediabetic State |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Enrollment | 3 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Emory University |
| Results posted | Yes |
2. Clinical Question
Population
The study concerned people with impaired glucose tolerance, described in the registry record as a form of prediabetes. The record lists an enrollment of 3. It does not provide further eligibility details or a description of the characteristics of those enrolled.
Intervention
One arm was assigned sitagliptin 100 mg daily by mouth together with pioglitazone 15 mg daily by mouth. A second arm was assigned sitagliptin 100 mg daily by mouth together with a pioglitazone placebo.
Comparator
The third arm received placebo for both sitagliptin and pioglitazone. The planned three-arm design therefore included the combination, sitagliptin with pioglitazone placebo, and placebo for both medicines. The registry record does not report arm-specific enrollment.
Primary question
The study aimed to determine whether sitagliptin, alone or combined with pioglitazone, would improve insulin secretion and insulin response in people with impaired glucose tolerance. The registered primary outcome was insulin secretion, but it was not measured because the study was prematurely terminated following unanticipated delays.
The clinical question has a treatment-comparison structure: the combination regimen could be compared with sitagliptin alone and with placebo, while the sitagliptin-only regimen could also be compared with placebo. These contrasts would address different parts of the stated aim. The comparison with sitagliptin alone would concern the addition of pioglitazone; the comparison with placebo would concern the active regimen against no active study medicine. However, the registry record contains no outcome measurements or analyses with which to estimate any of these contrasts.
Insulin secretion was intended to be assessed at baseline, 6 months, and 9 months, with the latter time point specified as occurring after a 3 month washout. These planned assessment times describe the outcome schedule, not observed data. The record does not report a measured treatment effect, an uncertainty interval, or a statistical test. As a result, the clinical question remains unanswered by the recorded outcome information.
3. Trial Design
The registry record classifies the study as randomized, quadruple-masked, and parallel. Random assignment is intended to make treatment groups comparable on average, while masking is intended to limit the influence of treatment knowledge on care, outcome measurement, and interpretation. These are general purposes of the design features; the record does not describe the specific randomization procedures or explain which roles were included in the quadruple masking.
The parallel structure assigned participants to one of three regimens: sitagliptin with pioglitazone, sitagliptin with pioglitazone placebo, or placebo for both medicines. The active combination was described as sitagliptin 100 mg daily by mouth plus pioglitazone 15 mg daily by mouth. The sitagliptin-only regimen used the same stated sitagliptin dose and a pioglitazone placebo. The placebo arm used placebo for both study medicines. Arm-specific enrollment and allocation details are not reported in the registry record.
The condition listed was prediabetic state, and the summary identifies impaired glucose tolerance as the relevant form of prediabetes. The primary outcome was insulin secretion, with planned measurements at baseline, 6 months, and 9 months after a 3 month washout. The registry record states that the outcome was not measured because the study was prematurely terminated due to unanticipated delays.
The record lists a start of 2011-07, primary completion of 2011-12, and a terminated status. It reports enrollment of 3. These details establish that the study did not produce the registered primary outcome data, but do not explain the delays beyond describing them as unanticipated. The phase is recorded as NA. Further protocol details, including eligibility criteria, study-site procedures, and the planned duration of treatment, 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.
Overall Study
| Stage | Sitagliptin + Pioglitazone PLACEBO | Sitagliptin + Pioglitazone | PLACEBO |
|---|---|---|---|
| STARTED | 0 | 0 | 0 |
| COMPLETED | 0 | 0 | 0 |
| NOT COMPLETED | 0 | 0 | 0 |
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 | Sitagliptin + Pioglitazone PLACEBO | Sitagliptin + Pioglitazone | PLACEBO | Total |
|---|---|---|---|---|---|
| Participants analysed | 0 | 0 | 0 | 0 | |
| Sex: Female, Male: Female | Participants | — | — | — | — |
| Sex: Female, Male: Male | Participants | — | — | — | — |
6. Primary Endpoint
The registry lists the primary outcome as Insulin Secretion, measured over baseline, 6 months, 9 months (after a 3 month washout).
| Element | Registry entry |
|---|---|
| Result type | N/A |
| Measure of spread | not applicable |
| Unit | not stated |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint was insulin secretion. In general, insulin secretion describes the release of insulin by the body in response to physiological demands, including changes in glucose. It is a relevant biological measure for a study focused on impaired glucose tolerance because it concerns one aspect of the body's regulation of blood glucose. The registry summary also refers to insulin response, but the registered primary outcome is specifically named insulin secretion.
The record planned assessments at baseline, 6 months, and 9 months after a 3 month washout. Repeated assessments can, in general, describe how a measure changes over time and whether patterns differ between assigned groups. A baseline assessment can provide a reference point for later measurements, while a later assessment after washout can address outcomes after the stated washout period. The registry record does not specify the measurement method, the endpoint scale, or how values at different time points were intended to be combined or compared.
Most importantly, the registry record says insulin secretion was not measured because the study was prematurely terminated due to unanticipated delays. Thus there are no reported endpoint values or treatment-effect estimates to interpret. The planned outcome and its rationale should not be confused with an observed result: this record describes what the study intended to assess, but does not show whether insulin secretion changed in any arm.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record does not report an estimate of the treatment effect on insulin secretion. Although insulin secretion was the primary outcome and assessment times were specified, the outcome was not measured after premature termination due to unanticipated delays. There is therefore no observed between-group contrast, no reported uncertainty interval, and no inferential test for the primary question.
The listed enrollment is 3, but that figure does not establish how participants were distributed among the three arms. Arm-specific counts are not reported in the registry record. With no outcome values and no reported arm counts, it is not possible to assess whether the planned comparisons were carried out or to characterize the precision that an estimate might have had. No treatment effect should be inferred from the study aim, the planned doses, or the existence of a randomized design.
The registry record includes adverse-event group entries with zero participants at risk and zero affected for serious and other events. These entries do not establish that no adverse events occurred among treated participants: the recorded at-risk counts are also zero, and the study does not report an analyzed safety population. Death counts are not reported. The safety entries consequently do not provide an interpretable estimate of event frequency or comparative safety.
In general, a confidence interval expresses the range of effect estimates compatible with the data under the method used, and its width conveys precision. No such interval is reported for the primary outcome. A p-value would address compatibility with a statistical null model under specified assumptions, not the size or clinical importance of an effect; no primary-outcome p-value is reported either. These concepts cannot fill the gap left by an outcome that was not measured.
The record identifies the trial as terminated and gives unanticipated delays as the reason the primary outcome was not measured. It does not describe any completed outcome analysis. A registry entry with a planned endpoint is not evidence of benefit, lack of benefit, or equivalence. The appropriate interpretation is that the registered clinical question has no reported answer in the outcome data for this study.
Planning illustrations described for the study are not a substitute for observed results. The stated planning assumptions use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the target effect specified as a fraction of the observed effect. Because the primary outcome was not measured, those assumptions cannot create an observed effect estimate. Any sample-size figure derived under such assumptions is a planning illustration, not observed power and not evidence about treatment efficacy.
9. Statistical Methods Explained
What does randomization contribute?
In general, randomization assigns participants to treatment groups by chance. Its purpose is to reduce systematic differences between groups at the point of assignment, including differences in known and unknown characteristics. It does not guarantee that groups will be identical, particularly in a small study, and it cannot compensate for an outcome that was never measured. The registry record identifies the design as randomized but does not describe the sequence-generation method, allocation concealment, or procedures used to implement assignment.
What does masking mean?
The study is recorded as quadruple-masked. In general, masking can reduce bias when expectations might affect treatment delivery, participant reports, outcome assessment, or analysis decisions. The record does not identify the roles covered by masking or describe how masking was maintained. The classification alone should therefore not be taken as proof that every potential source of bias was controlled.
How should the parallel arms be understood?
In a parallel design, assigned groups receive different regimens and are followed as separate groups. This makes between-group comparisons the natural way to address a treatment question. The three registered regimens offer more than one possible contrast: the combination versus sitagliptin with pioglitazone placebo, the combination versus placebo for both medicines, and sitagliptin with pioglitazone placebo versus placebo for both. These contrasts correspond to different questions and should not be treated as interchangeable.
The registry record does not report a prespecified hierarchy or analysis plan for those possible comparisons. It also does not give an arm-specific allocation count. Without outcome values and arm counts, no treatment contrast can be reconstructed from the registry record.
What would a repeated endpoint measure?
The intended endpoint, insulin secretion, was scheduled at baseline, 6 months, and 9 months after a 3 month washout. In general, repeated measurements may be analyzed to compare changes over time, to compare groups at a specified time, or to model both group and time patterns. Those approaches answer different questions and can depend on assumptions about correlation among measurements from the same participant. The registry record does not specify which approach was planned, how the endpoint was measured, or how missing scheduled measurements would be handled. The endpoint was not measured, so no longitudinal analysis is reported.
What do significance levels and planning assumptions mean?
The planning description specifies a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. In general, a significance level is a threshold used to assess data against a statistical null hypothesis under a chosen testing procedure. It is not the probability that a treatment works, nor does it describe the precision of an estimate by itself. A normal approximation is a mathematical simplification that can be used for certain planning calculations; its suitability depends on the outcome and design.
The planning description also expresses the target effect as a fraction of an observed effect. Such an illustration is conditional on an effect estimate and selected assumptions. It does not show that the trial was designed using those assumptions, and it is not observed power. The registry record says the primary outcome was not measured, so the study does not report an observed primary effect to interpret. Numerical planning estimates and a trial-specific sample-size justification are not reported in the registry record.
Why can over-dispersion matter in general?
For outcomes that are counts, over-dispersion means the observed variability is greater than a model assumes. If a model understates variability, standard errors may be too small and intervals may be too narrow. This is a general statistical concern, not a reported feature of this study. The primary endpoint is described as insulin secretion, with no measurement scale or distributional information reported; there is no basis in the registry record to characterize its dispersion or select a particular model.
What can be concluded from the posted analysis information?
The record reports no analyses. It therefore supplies no fitted model, test statistic, estimate, interval, or analysis population for the primary outcome. Statistical methods can explain what these quantities would mean if reported, but cannot replace the absent outcome data. The appropriate reading is descriptive: the design and intended outcome are documented, while a statistical result for the primary question is not.
10. Multiplicity, Missing Data and Analysis Populations
The registered primary outcome had planned measurements at baseline, 6 months, and 9 months after a 3 month washout. Multiple time points can create more than one potential comparison. In general, when several hypotheses are tested, the chance of obtaining at least one apparently notable result can increase unless the analysis plan addresses multiplicity. A prespecified primary time point, an ordered testing strategy, or an adjustment method are examples of general approaches, but the registry record does not report whether any such approach was planned.
The record also lists three treatment arms, which allow several possible pairwise contrasts. It does not report a multiplicity strategy or identify a primary treatment contrast. The registry outcome information was not reduced to one comparison, and no analyses are reported. Because the primary endpoint was not measured, there are no reported tests across time points or arms to evaluate.
Missing-data handling is not reported in the registry record. That absence should be distinguished from the statement that the outcome was not measured due to premature termination: the record does not describe whether any partial measurements existed, why they might be missing at an individual level, or whether any imputation or sensitivity analysis was considered. The analysis population is also not reported. No intention-to-treat, per-protocol, or safety population is identified in the outcome information.
In general, missing-data methods rely on assumptions about why observations are absent and how missingness relates to measured or unmeasured outcomes. Those assumptions cannot be assessed from the registry record. The absence of an analysis plan, multiplicity details, and analysis-population definitions limits interpretation, but it does not justify guessing which methods were used.
11. What the Result Does, and Does Not, Mean
The record establishes that a randomized, quadruple-masked, parallel study was registered to examine sitagliptin, with or without pioglitazone, in people with impaired glucose tolerance. It identifies insulin secretion as the primary outcome and gives intended assessment times. It also states that the endpoint was not measured because the study was prematurely terminated due to unanticipated delays. These registry entries describe the study's aim and its incomplete outcome record, not a treatment result.
The study does not show that either active regimen improved insulin secretion, nor does it show that either regimen failed to do so. It also does not establish equivalence between regimens or between an active regimen and placebo. Those conclusions require observed outcome data and an analysis capable of estimating the relevant contrast. Neither is reported for the primary endpoint.
The enrollment entry of 3 is not a result and does not tell how many participants were assigned to each arm or contributed measurements. The adverse-event entries, with zero listed at risk as well as zero affected, do not demonstrate the absence of adverse events among participants exposed to treatment. Event-rate comparisons cannot be interpreted from those entries.
In general, a study can be statistically uninformative about its target question when the outcome is unavailable, regardless of whether the planned design could have supported a comparison under other circumstances. A planned time point is not an observed assessment, and a planning calculation is not a treatment-effect estimate. The registry record supports a conclusion about the status of the evidence: the registered primary outcome has no reported measurement or analysis. It does not support a conclusion about the clinical effect of the medicines.
12. Why This Trial Matters Statistically
This study illustrates a central distinction in clinical-trial interpretation: a clearly stated research question and design do not guarantee an interpretable result. The record specifies randomization, masking, three treatment regimens, and an insulin-secretion outcome, but also states that the outcome was not measured after premature termination. The statistical question therefore remains open in this record.
The distinction matters because readers may otherwise mistake a planned endpoint or treatment regimen for evidence about an effect. Statistical inference requires observed data for the outcome being compared. Without those data, an estimate cannot be calculated or judged for precision, and a significance test cannot address the registered question. The trial's classification as randomized does not change that limitation.
The record also demonstrates why arm-level information and a prespecified analysis plan are important for interpreting multi-arm studies. Different contrasts answer different questions, and repeated outcome times can introduce further analytical choices. The registry record does not report a multiplicity strategy or analysis details. Since there are no outcome analyses, however, these issues are constraints on what can be interpreted, not evidence that a particular analytical choice changed a result.
Finally, termination and missing outcome measurement should be visible in the statistical account of a trial. They define the boundary between the intended design and the evidence actually recorded. For this study, the main statistical lesson is not a numerical treatment finding; it is the need to distinguish planned measurement and planning assumptions from analyzed clinical outcomes.
13. 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 limitation is that insulin secretion, the registered primary outcome, was not measured. The registry record attributes this to premature termination due to unanticipated delays. Consequently, there is no reported primary-outcome estimate, uncertainty interval, or statistical test.
The record lists enrollment of 3 but does not report participant counts by arm, baseline characteristics, adherence, or follow-up completion. It therefore does not establish how many participants were assigned to each regimen or whether any planned comparisons were feasible. The details of randomization and the roles covered by quadruple masking are not described.
Outcome measurement procedures and the endpoint scale are not reported. Although baseline, 6 months, and 9 months after a 3 month washout are listed as time points, the analysis method for repeated measurements is not given. The registry record also does not report how missing observations, multiplicity, or analysis populations would be handled.
The safety entries list zero participants at risk and zero affected in the recorded groups, while deaths are not reported. These entries do not support an assessment of safety among treated participants. Further details about adverse-event collection and the safety population are not reported in the registry record.
The registry record does not give a trial-specific sample-size justification or numerical planning result. It describes general planning assumptions, but these cannot be treated as the trial's design rationale or as evidence from observed outcomes. Together, these limitations mean the record describes the intended study but does not provide an analyzable account of its primary clinical outcome.
14. Related Tutorials
15. Related Statistical Calculators
16. Sources
- ClinicalTrials.gov: DPP-4 Inhibition and TZD for DM Prevention — NCT01006018.
- Registry record last updated 2013-12-05; status last verified 2013-11.
- Results first posted 2013-06-07.
- 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.
17. Record Summary
NCT01006018 is a randomized, quadruple-masked, parallel study listed under the title DPP-4 Inhibition and TZD for DM Prevention. It concerned impaired glucose tolerance and compared a planned combination of sitagliptin and pioglitazone with sitagliptin plus pioglitazone placebo and placebo for both medicines. The record lists enrollment of 3 and a terminated status.
Insulin secretion was the primary outcome, with planned assessment at baseline, 6 months, and 9 months after a 3 month washout. The registry record states that this outcome was not measured because the study was prematurely terminated due to unanticipated delays. No analyses are reported, so no primary treatment-effect estimate or inferential result is available. The safety entries do not establish event rates because the listed at-risk counts are zero. The registry record therefore documents the intended question and design, but does not report an outcome-based answer.