Dipeptidyl Peptidase-4 Inhibition in Psoriasis Patients With Diabetes (DIP): A Randomized Clinical Trial.: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Dipeptidyl Peptidase-4 Inhibition in Psoriasis Patients With Diabetes (DIP): A Randomized Clinical Trial. (NCT01991197), a randomized phase 2 study. The primary outcome is The Change in the Psoriasis Area and Severity Index (PASI) From Baseline to 16 Weeks in Psoriasis Patients With Type 2 Diabetes Treated With Sitagliptin Compared to Patients Treated With Gliclazide..

2. Clinical Question

The question concerns a psoriasis severity score in people who also have type 2 diabetes, rather than a comparison of diabetes outcomes. PASI combines assessment of psoriasis extent and severity; the registry describes its scale as ranging from 0 to 72, with higher scores indicating worse outcome. The primary outcome is change from baseline, but the posted arm results give medians and interquartile ranges without separate baseline values or median changes. Therefore, the values 9.5 and 9.4 should be read as the reported group summaries for the primary outcome, not as a treatment effect or as the amount of change from baseline.

The comparison belongs to the double-blind portion of a study that also included an open-label period. After the 16-week comparison, the registry summary says all participants would receive sitagliptin for a further 16 weeks. The primary question is specifically the between-group comparison at 16 weeks; it is not a comparison of continued randomized treatments through the later period.

3. Trial Design

This was a randomized phase 2 trial with a crossover model recorded in the registry. The enrollment was 20, and the trial status is completed. The stated conditions were psoriasis and type 2 diabetes mellitus. University College Dublin is listed as sponsor. Recruitment was described as taking place at two centres.

The registry summary describes a four-week washout period, followed by prospective follow-up for 36 weeks. Participants were to be stratified by centre, psoriasis severity, and obesity status, then randomly allocated to Arm A or Arm B. During the first 16 weeks, one arm received sitagliptin with matched placebo for gliclazide, while the other received gliclazide with matched placebo for sitagliptin. The summary then describes an open-label phase in which all participants received sitagliptin for a further 16 weeks. These details describe the planned sequence; the registry record does not report participant-level treatment histories.

The masking field identifies quadruple masking, while the narrative specifically says participants and investigators would be unaware of the allocated arm. The registry record does not explain which additional roles were masked. It is therefore possible to report the recorded masking designation and the narrative description, but not to identify all masked parties from the registry record.

The primary outcome was assessed at 16 weeks, at the end of the initial treatment comparison. The registry reports 9 participants in the sitagliptin group and 11 in the gliclazide group for this outcome. These outcome counts add up to the stated enrollment, but the registry record does not explain whether all randomized participants were included in the analysis or how the analysis population was defined. The term crossover appears in the design metadata; the posted primary analysis is a comparison between the two groups at 16 weeks, and the registry record does not give enough detail to reconstruct a within-participant crossover analysis.

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.

6. Primary Endpoint

The registry lists the primary outcome as The Change in the Psoriasis Area and Severity Index (PASI) From Baseline to 16 Weeks in Psoriasis Patients With Type 2 Diabetes Treated With Sitagliptin Compared to Patients Treated With Gliclazide., measured over 16 weeks.

The primary endpoint is change in the Psoriasis Area and Severity Index (PASI) from baseline to 16 weeks. PASI is a score of psoriasis extent and severity. The registry describes a scale from 0 to 72, where higher scores indicate worse outcome. Change over time is used to describe whether the measured severity score shifted during the treatment period, while comparison between the randomized groups can help distinguish the observed change under one assigned regimen from change under the comparator regimen.

The registry identifies the endpoint summary as a median with interquartile range. A median describes the middle of the observed values after they are ordered. The interquartile range describes the middle half of those values and gives information about their spread. These summaries are useful when data may be skewed or when a small number of observations could strongly affect a mean. They do not, by themselves, describe the uncertainty around the between-group difference.

The posted group summaries are 9.5, with an interquartile range of 7.8 to 9.6, for sitagliptin, and 9.4, with an interquartile range of 7.6 to 12.0, for gliclazide. The registry’s outcome description labels the endpoint as change from baseline, but does not provide separate baseline and follow-up summaries or a numerical between-group estimate of change. That limits interpretation: the reported medians should not be treated as the magnitude of change from baseline. The registry record also does not report a confidence interval for the endpoint comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the Mann-Whitney test compare?

The registry reports a Wilcoxon (Mann-Whitney) test, a rank-based method for comparing observations between two groups. Rather than requiring the analysis to compare group means, the method orders the observations and uses their ranks to assess whether values in one group tend to be higher or lower than values in the other. It is commonly used when a continuous outcome is summarized by medians and interquartile ranges or when a normal-distribution assumption is doubtful.

In general, a rank-based test does not automatically test a difference between medians in every situation. Its interpretation as a location or median comparison is most straightforward when the groups have similarly shaped distributions. If distributions differ in spread or shape, the test can also respond to those differences. The registry does not provide distribution plots or individual scores, so the shape of the PASI distributions cannot be assessed from the reported summaries alone.

The posted test statistic is U = 43.5 and the p-value is 0.648. The p-value does not measure the size of the treatment contrast. A small p-value can arise from a sufficiently consistent difference relative to the observed variation, while a larger p-value indicates that the data do not provide strong evidence against the null comparison under the test. A larger p-value does not establish that the groups are identical.

Why are medians and interquartile ranges reported?

A median is the central ordered observation, and the interquartile range gives the bounds for the middle half of the distribution. These summaries are less affected by extreme values than a mean and standard deviation. They can be informative for outcomes whose distributions are skewed. Here, the registry specifies median and interquartile range as the summary form, with group medians of 9.5 and 9.4. The interquartile ranges are 7.8 to 9.6 for sitagliptin and 7.6 to 12.0 for gliclazide.

In general, overlap between interquartile ranges is not a formal test of a treatment difference. Nor does their overlap determine whether a difference is clinically important. A confidence interval for a relevant between-group effect would help show the range of values compatible with the data. No confidence interval is reported for the primary comparison in the registry.

What does the p-value say about uncertainty?

In general, a p-value is calculated under a specified null hypothesis and test procedure. It describes how unusual a test statistic at least as inconsistent with the null as the observed statistic would be, if the null model and assumptions held. It does not give the probability that the treatment groups are equal, and it does not state the chance that a finding is due to chance in everyday terms. The reported p = 0.648 should be read as the result of the listed rank-based comparison, not as a measure of benefit, harm, or equivalence.

A confidence interval answers a related but different question by expressing uncertainty around an estimated effect. Its width conveys precision, and its position relative to a meaningful reference value helps assess which effects remain compatible with the data. The registry does not report a confidence interval or a numerical between-group estimate of change, so the size and precision of the primary contrast cannot be fully assessed from the posted result.

What do the sample-size illustrations represent?

The registry record does not report a trial-specific sample-size justification. Planning illustrations use the observed effect as a basis, with a two-sided significance level of 0.05, equal allocation, and a normal approximation; the intended effect is specified as a fraction of the observed effect. Such values are planning illustrations, not observed trial results and not observed power. The specific planning counts are not reported in the registry record.

In general, a planning calculation asks how many participants might be needed to detect a prespecified effect under stated assumptions. It does not show the probability that a completed study has demonstrated an effect. An illustration based on an observed effect is especially dependent on that estimate, which can be unstable in a small study. It should not be read as evidence that the trial was adequately powered or that a non-significant result rules out a meaningful effect.

What is not specified about the analysis?

The registry names the test and reports its statistic and p-value, but further details of the statistical plan are not reported in the registry record. This includes whether the analysis adjusted for the stratification factors, how the change score was calculated, how missing outcome data were handled, whether the analysis followed randomized assignment, and whether any sensitivity analyses were conducted. It would be inappropriate to infer those choices from the test name alone.

In general, missing data can affect both the size and reliability of an estimate, especially when the likelihood of missingness is related to outcome or treatment. The registry gives group counts for the primary outcome but does not describe reasons for missingness or an imputation method. The reported comparison can be described as posted, while its handling of missing data and analysis population remains uncertain.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a Wilcoxon (Mann-Whitney) test, with U value 43.5 and p = 0.648. It also reports the group medians and interquartile ranges. These are the posted results to use when describing the sponsor’s analysis. The registry does not report a confidence interval or a separate numerical estimate of the between-group change.

No discrepancy is flagged between the sponsor’s reported analysis and a simple unadjusted recalculation. The posted analysis therefore remains the relevant result to describe. A recalculation from summary statistics, when available, would be a separate unadjusted analysis and would not replace the sponsor’s analysis or recover details that depend on individual participant data. In this case, a recalculated effect estimate is not reported in the registry record.

The endpoint wording and summary values leave an interpretive limitation: the outcome is described as change from baseline, but the record does not give separate baseline and follow-up summaries or spell out how the reported medians were constructed. The sponsor’s p-value can be reported as the test result, but it cannot resolve that ambiguity or supply an unreported effect estimate. Accordingly, the statistical reading should distinguish the reported test result from claims about the amount of change or the precision of a treatment difference.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 16 outcome measures and 13 posted secondary outcomes. The secondary outcomes cover several domains, including adverse events, quality of life, other efficacy endpoints, inflammatory markers, PASI at a later time, weight, and systolic blood pressure. The primary endpoint is PASI change at 16 weeks. Considering multiple outcomes can create more opportunities for apparently unusual results, especially when many tests are conducted.

The registry record does not report whether the statistical plan adjusted for multiple comparisons, identified a hierarchy among secondary outcomes, or designated any analyses as exploratory. It also does not report whether interim analyses were planned or performed. The results for secondary outcomes should therefore be interpreted as separate posted findings, without assuming that a multiplicity adjustment was or was not applied.

The primary outcome results list 9 participants in the sitagliptin group and 11 in the gliclazide group, matching the stated enrollment of 20. The registry record does not describe the analysis population, confirm how participants were handled after randomization, or explain whether all randomized participants were included under their assigned group. The reasons for any missing outcome data, if any, and the methods used to handle missingness are not reported in the registry record.

In general, missing-data methods matter because excluding participants can alter the comparison if outcome availability is related to treatment assignment, participant characteristics, or outcome values. An intention-to-treat analysis, a per-protocol analysis, and an analysis limited to participants with observed outcomes answer different questions and rely on different assumptions. The registry does not identify which approach was used for the primary endpoint. It also does not report sensitivity analyses to test how alternative assumptions about missing outcomes might affect the findings.

These omissions do not show that the analysis was inappropriate; they limit what readers can verify from the registry results. The reported p-value describes the named test, but the record does not provide enough analysis-plan detail to assess multiplicity control, the handling of missing data, or the population used beyond the posted group counts.

12. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: vasovagal episode leading to anoxic seizure (11.1%), clinically isolated syndrome (9.1%), osteomyelitis (9.1%).

For serious adverse events, the registry reports 1 participant affected among 9 at risk in the sitagliptin group and 2 among 11 at risk in the gliclazide group. It reports 0 deaths in each group. For other adverse events, the counts are 6 among 9 at risk with sitagliptin and 10 among 11 at risk with gliclazide. These are counts from the registry record and should be read with the group sizes in mind.

The listed serious-event terms are vasovagal episode leading to anoxic seizure at 11.1%, clinically isolated syndrome at 9.1%, and osteomyelitis at 9.1%. The registry does not specify an arm beside each of these terms in the posted list, so they should not be assigned to a treatment group based on the list alone. The record does not provide further detail about event timing, duration, or adjudication in the registry record reported.

In general, safety counts from a small trial are descriptive and can be highly uncertain. A difference in counts does not by itself establish a treatment-related cause, and a zero count does not establish that an event cannot occur. The registry’s adverse-event figures do not support a precise comparative safety estimate, and they should not be interpreted as proof that one regimen is safer. The record does not report a formal statistical comparison for these counts.

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

The posted primary result describes a small randomized comparison of sitagliptin and gliclazide in participants with psoriasis and type 2 diabetes. At 16 weeks, the reported median PASI values were 9.5 and 9.4, respectively, and the Wilcoxon (Mann-Whitney) test gave p = 0.648. The result does not provide evidence of a statistically significant difference in the reported rank-based comparison.

That finding does not mean the treatments are equivalent. Equivalence requires a defined margin and an analysis designed to assess whether the plausible treatment difference lies within that margin. The registry record does not report such a margin or an equivalence analysis. A non-significant test also does not show that there was no change in either group; the posted comparison is between groups, and the record does not provide the baseline and follow-up summaries needed to describe change separately for each group.

The result does not establish how large a difference in change is compatible with the data because a confidence interval and a numerical between-group estimate are not reported. The sample was small, and the posted outcome counts were 9 and 11. As a general statistical point, limited sample sizes often leave considerable uncertainty, even when the observed group summaries are close. The p-value alone cannot determine whether effects that matter to patients remain compatible with the evidence.

The endpoint is a measure of psoriasis severity, not a measure of diabetes control. The trial’s later open-label period, when all participants were described as receiving sitagliptin, does not offer the same randomized contrast as the primary 16-week comparison. The registry record also does not fully describe the analysis population, missing-data handling, or multiplicity strategy, so those aspects cannot be used to strengthen or weaken the result beyond noting the uncertainty.

The appropriate reading is narrow: the registry reports similar group medians and a non-significant rank-based test for the primary comparison at 16 weeks. This alone neither proves a treatment effect nor rules one out, and it does not support a conclusion of equivalence, clinical importance, or comparative safety.

14. Why This Trial Matters Statistically

This trial illustrates a common challenge in clinical research: a clear clinical question can be paired with a result that is difficult to interpret fully from summary statistics alone. The primary outcome is a change in psoriasis severity, yet the record provides group medians and interquartile ranges without separate baseline and follow-up summaries or a numerical estimate of the difference in change. That distinction matters because a group’s score and its change from baseline are not interchangeable descriptions.

The reported test is rank-based and fits the use of median summaries, but a p-value does not describe effect size or precision. The value p = 0.648 is not evidence that the groups are equivalent. Without a confidence interval or an explicit equivalence framework, readers cannot determine which treatment differences remain compatible with the data. This is a useful reminder that the absence of a statistically significant result is not the same as affirmative evidence of no meaningful difference.

The trial’s randomized design and active comparator make the initial treatment comparison relevant to the stated question. At the same time, the modest enrollment and the small outcome groups limit the detail that can be learned from the result. The later open-label period answers a different kind of question because the registry summary says all participants then received sitagliptin.

Finally, the record shows why design labels and analysis details need to be read together. It records a crossover model, describes randomized groups during the first 16 weeks, and reports a two-group rank test at that time point. Further information about the analysis population, missing outcomes, and multiplicity is not reported in the registry record. Those gaps do not invalidate the posted result, but they define the boundaries of what can be concluded from it.

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 primary result is based on 9 participants in the sitagliptin group and 11 in the gliclazide group. Although the registry states enrollment of 20, the record does not explain how the primary analysis population was defined or describe participant flow through randomization and follow-up. It does not report whether there were missing outcome measurements or how such measurements were handled.

The endpoint is identified as change in PASI from baseline to 16 weeks, but the reported arm summaries do not provide baseline values, follow-up values, or a distinct between-group estimate of change. The record does not explain how the posted medians and interquartile ranges relate to the change endpoint. No confidence interval is reported, limiting assessment of precision.

The statistical description names a Wilcoxon (Mann-Whitney) test and reports its statistic and p-value, but does not give additional details about model adjustments, the prespecified analysis plan, or sensitivity analyses. The record also does not report a sample-size justification, interim analyses, or a plan for handling multiple outcome comparisons. Those omissions should be treated as limitations of the registry information, not as evidence that the choices were absent from the full study materials.

The design metadata records a crossover model, while the narrative describes two randomized groups during the first 16 weeks and an open-label phase in which all participants receive sitagliptin. The posted primary analysis is a group comparison at 16 weeks. The registry record does not explain how the crossover label relates to the reported primary analysis or provide enough detail to reconstruct later-period comparisons.

Masking is recorded as quadruple, while the narrative specifically identifies participants and investigators as unaware of allocation. The additional masked roles are not reported. The trial also recruited at two centres, but centre-specific outcome results are not reported. With a small sample and limited analysis detail, the registry results give a restricted view of the primary comparison and should not be generalized beyond the recorded population and endpoint without additional evidence.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Dipeptidyl Peptidase-4 Inhibition in Psoriasis Patients With Diabetes (DIP): A Randomized Clinical Trial. — NCT01991197.
• Registry record last updated 2020-05-14; status last verified 2020-05.
• Results first posted 2020-05-14.
• 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

NCT01991197 is a completed randomized phase 2 trial sponsored by University College Dublin. It enrolled 20 participants with psoriasis and type 2 diabetes mellitus. The registry describes a 16-week comparison of sitagliptin with gliclazide, followed by an open-label period in which all participants received sitagliptin.

The primary endpoint was PASI change from baseline to 16 weeks. The registry reports median values of 9.5 for sitagliptin and 9.4 for gliclazide, with interquartile ranges of 7.8 to 9.6 and 7.6 to 12.0. A Wilcoxon (Mann-Whitney) test reported U value 43.5 and p = 0.648. The outcome counts were 9 and 11 participants. No confidence interval or numerical between-group estimate of change is reported.

The record also reports adverse-event counts and multiple secondary outcomes. Details about missing-data handling, analysis populations, multiplicity, and the relation between the crossover design label and the primary comparison are not reported in the registry record.