Effects of Sitagliptin on Postprandial Lipemia in Men With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effects of Sitagliptin on Postprandial Lipemia in Men With Type 2 Diabetes (NCT00660075), a randomized phase 3 study. The primary outcome is Measurement of the Area Under the Curve of Plasma Triglycerides (TG) Levels During Postprandial Period (Time 0,2,4,6,8 Hours).

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, quadruple-masked, placebo-controlled crossover study. Its official title describes an evaluation of sitagliptin's effects on postprandial plasma lipoprotein concentrations in men with type 2 diabetes. The registry lists Laval University as sponsor and the trial status as completed. The recorded start is 2008-02 and primary completion is 2009-01.

The intervention arms were sitagliptin 100 mg/d for 6 weeks and placebo for 6 weeks. In a crossover design, participants receive treatment in more than one study period, allowing comparison across treatment conditions within the study population. That general design can make within-participant contrasts informative, but interpretation depends on details such as treatment sequence, washout, and how period effects are handled. Those details are not reported in the registry record.

The outcome description says all subjects involved were analyzed per protocol and that the placebo phase was compared with the sitagliptin phase. The primary outcome was measured at the end of the two 6-week interventions. The registry lists 36 enrolled participants and reports 36 observations for each treatment phase in the primary outcome results. These are the posted phase-specific outcome counts; they should not be read as separate parallel treatment groups.

The registry record does not report the original sample-size justification, the planned statistical model, treatment sequence allocation details, washout duration, or interim analyses. It also does not post secondary outcome results or analysis records. These omissions limit reconstruction of the protocol and assessment of whether every planned analysis is represented.

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: 2-weeks run-in period. 6-weeks treatment with sitagliptin 100mg/d or placebo. 4-weeks washout period. 6-weeks treatment with sitagliptin or placebo.

First Intervention (Week 1 to Week 6)

Washout Period (Week 7 to Week 10)

Second Intervention (Week 11 to Week 16

Analysis population for the primary outcome, as defined in the registry: We analyzed all the subjects involved in the study. The analysis was per protocol. We compared data from the placebo phase with the sitagliptin phase.

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 Measurement of the Area Under the Curve of Plasma Triglycerides (TG) Levels During Postprandial Period (Time 0,2,4,6,8 Hours), measured over At the end of the two 6-week interventions.

The primary endpoint was the area under the curve of plasma triglyceride levels during the postprandial period, measured at time 0, 2, 4, 6, and 8 hours. Its unit is mmol*h/L, and the registry identifies the reported measure as a mean with standard deviation. The outcome was assessed at the end of both 6-week intervention periods.

In general, an area under the curve combines measurements across a specified time window into a summary of overall exposure during that window. For postprandial lipemia, this approach captures the pattern of triglyceride levels after a meal across the scheduled sampling times rather than relying on one measurement alone. It is therefore a measure of post-meal triglyceride exposure over the stated observation period. It does not, by itself, describe the timing of a peak or show the individual measurements at each time point.

The registry's primary outcome is specifically triglyceride exposure. Although the study title and background refer to plasma lipoproteins more broadly, the posted primary result should not be treated as a result for every lipoprotein measure or for cardiovascular outcomes. The record does not provide a detailed description of meal composition, sampling procedures, or the endpoint calculation beyond the listed time points and area-under-the-curve definition.

7. Results Posted in the Registry

• Sitagliptin 100 mg/d: mean 27.10 versus 30.00 on control; difference -2.90 (95% CI -9.26 to 3.46), standardized effect d = -0.21; z-test p = 0.372.

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation starts with identifying the estimand: what comparison the result is intended to describe. Here, the primary result is a comparison of mean postprandial triglyceride area under the curve between the placebo and sitagliptin phases. The reported mean difference is -2.9 mmol*h/L, calculated in the direction sitagliptin relative to placebo. The result is a continuous-outcome contrast, not a comparison of event counts or proportions.

What does a mean difference describe?

In general, a mean difference expresses how far apart the average outcome values are on the outcome's original measurement scale. It retains the clinical measurement units, which makes the estimate interpretable in mmol*h/L. The phase-specific means were 30.0 mmol*h/L for placebo and 27.1 mmol*h/L for sitagliptin. The posted contrast is -2.9 mmol*h/L. No further transformation is needed to understand the direction: a negative contrast means the reported mean was lower in the sitagliptin phase.

A mean is a summary and does not show the distribution of individual values. Two phases can have similar means while participant-level changes vary considerably. The standard deviations of 14.5 mmol*h/L and 13.0 mmol*h/L indicate substantial variation around the respective phase means, but those values do not reveal how each participant's two measurements relate to one another. In general, a crossover analysis can use that within-participant relationship, so phase-specific standard deviations alone are not enough to reproduce its uncertainty calculation.

Why does the crossover structure matter?

In general, a crossover trial compares treatments given in different periods, often allowing a participant to serve as their own reference. This can control for stable differences between participants. The validity and precision of the comparison can also depend on period effects, treatment sequence, carryover, and the analysis model. The registry identifies the model as crossover and says the placebo phase was compared with the sitagliptin phase, but does not report sequence, washout, period adjustment, or carryover assessment. These features should not be assumed.

The outcome description says the analysis was per protocol and included all subjects involved in the study. That wording does not explain how protocol deviations, incomplete treatment periods, or missing outcome measurements were handled. The posted count is 36 in each phase. The registry does not provide a participant flow or reasons for exclusion, so the meaning of the per-protocol population cannot be fully assessed.

How should the p-value be read?

In general, a p-value quantifies the compatibility of the observed data with a specified null hypothesis under the assumptions of a statistical test. The reported unadjusted value is 0.372. It is not an effect size and does not tell how large or important the difference is. It also does not establish that there is no difference. The specific test, model specification, and any treatment of the crossover pairing are not reported in the registry record. Accordingly, the p-value can be quoted as posted, but its calculation cannot be verified from the summary results.

The confidence interval supplies information about uncertainty on the outcome scale. It extends from -9.262 to 3.462 mmol*h/L. The interval includes zero, indicating that the posted data do not rule out a zero mean difference under the method used to form the interval. It also includes values on both sides of zero, illustrating uncertainty about direction as well as magnitude. A confidence interval is not a guarantee that the true effect lies within its endpoints; in general, its repeated-sampling interpretation depends on the statistical method and assumptions.

Why can variability and over-dispersion matter?

In general, high variability relative to the size of a mean difference can make an estimate less precise. In count outcomes, over-dispersion means the data vary more than a simple model expects, which can make standard errors too small if it is ignored. This primary endpoint is continuous, not a count, so over-dispersion is not the specific issue indicated by this result. For this endpoint, the available summaries show wide variation in phase-specific values and a confidence interval that is broad relative to the point estimate. The registry does not report diagnostics or a distributional model.

The posted comparison is not marked as recalculated, and the discrepancy flag is false. The reported estimate and inferential values are therefore presented as the registry's result. No sponsor analysis details are included beyond the posted results. A recalculation from summary statistics would not be a substitute for the original crossover analysis, particularly without the within-participant covariance.

Finally, the registry has no posted secondary analyses. The absence of these results does not establish that no other outcomes were planned or measured. It means that interpretation of the posted statistical findings is limited to the single primary outcome result described in the record.

10. Confidence Intervals and Precision

The reported confidence interval for the mean difference extends from -9.262 to 3.462 mmol*h/L. It describes uncertainty around the estimated contrast of -2.9 mmol*h/L, with sitagliptin compared against placebo. The interval is expressed on the same scale as the endpoint, mmol*h/L, so its endpoints can be understood as possible values for the mean difference under the interval's statistical assumptions.

Because the interval crosses zero, it is compatible with no mean difference as well as differences in either direction. Its lower endpoint represents a larger negative contrast than the point estimate, while its upper endpoint is positive. This breadth means the estimate is not precise enough to narrowly identify the direction and size of the contrast. The unadjusted p-value of 0.372 is consistent with the fact that the interval includes zero, although the registry does not report the exact interval method or test used.

In general, a confidence interval is generated by a procedure that would cover the target parameter at a stated rate over repeated samples, provided its assumptions hold. It is not a probability statement that the true effect is within these particular endpoints. The registry record does not report the confidence level explicitly, so no confidence level should be inferred from the record itself.

For a crossover design, precision can depend on how closely a participant's outcomes track across the treatment phases. The phase-specific standard deviations are 14.5 mmol*h/L for placebo and 13.0 mmol*h/L for sitagliptin, but those summaries do not give the within-participant covariance. The posted confidence interval is the relevant uncertainty summary; its calculation cannot be independently rebuilt from the registry's phase means and standard deviations alone.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total event counts. They are illustrations based on the observed effect, not the trial's original design justification. The registry record does not report the original sample-size assumptions, so the planning estimates should not be treated as evidence about how the trial was designed.

The listed assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect. For planning for 100% of the observed effect, the estimates are 354 participants per arm for 80% power and 474 participants per arm for 90% power. For planning for 75% of the observed effect, the estimates are 630 participants per arm for 80% power and 843 participants per arm for 90% power. For planning for 50% of the observed effect, the estimates are 1,416 participants per arm for 80% power and 1,896 participants per arm for 90% power.

In plain terms, the 80% power column gives the estimated participants per arm for a design with an 80% chance of detecting the specified effect under the planning assumptions, if that effect and assumptions are correct. The 90% power column gives the corresponding estimate for a design with a 90% chance. Smaller effects need more participants to distinguish them from random variation with the same planned power. These are planning illustrations, not observed power for this completed trial.

The calculation assumptions describe a two-arm, equal-allocation illustration. The clinical trial itself is a crossover study, so the planning figures should not be read as a reconstruction of its design or as a direct account of its paired analysis. The registry does not report the original sample-size calculation.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -2.900; pooled standard deviation σ = 13.770.
2. n = 2 × (1.96 + 0.842)² × 13.770² / -2.900² = 353.9, rounded up to 354 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -2.9                      # observed difference in means
sigma <- 13.7704393539204                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(660075)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 354 (80% power), 474 (90% power)
 75% of observed effect: n per arm = 630 (80% power), 843 (90% power)
 50% of observed effect: n per arm = 1,416 (80% power), 1,896 (90% power)
power.t.test: 354.9 
Simulated power at n = 354 per arm: 0.800 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry posts one primary outcome measure and no secondary outcome results. It also lists no analysis records. The primary result is the mean area under the curve of plasma triglycerides during the postprandial period. The absence of posted secondary results limits the range of questions that can be assessed from the registry record.

Multiplicity arises in general when several outcomes, time points, comparisons, or analyses are examined. If many inferential tests are conducted, the chance of obtaining at least one apparently notable result can increase unless the analysis plan accounts for that. The record does not describe whether multiplicity adjustment was planned or used. It also does not establish whether additional endpoints or analyses were measured but not posted. The primary p-value is identified as unadjusted, so it should be read as an unadjusted result rather than as evidence that multiplicity was handled.

The outcome description states that the analysis was per protocol and included all subjects involved in the study. It does not detail how participants were classified for that population, how protocol deviations were handled, or whether any measurements were missing. No missing-data method, imputation procedure, or sensitivity analysis is reported in the registry record. The posted phase-specific outcome counts are 36 for placebo and 36 for sitagliptin, but these counts do not explain the handling of incomplete observations or the participant flow.

For a crossover trial, analysis populations and missingness can be especially relevant because a within-participant comparison depends on observations across treatment periods. The registry does not say whether participants with data in only one phase contributed to any analysis, or how period-specific omissions were treated. These details are not reported in the registry record, and no assumptions about them should be made.

The protocol's statistical analysis plan, the trial's own sample-size justification, interim analyses, and adjustments for multiple comparisons are not reported in the registry record. Without these details, it is not possible to determine whether the posted result reflects all prespecified analyses or how the primary test was positioned among any additional analyses.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The serious adverse event display lists 18 participants at risk for sitagliptin and 18 at risk for placebo. It reports 0 serious adverse events in each treatment category. These are the registry's serious-event counts, and the record does not provide serious adverse event terms.

The display also lists 18 participants at risk for other adverse events in each treatment category and reports 0 other adverse events in each. The registry's death entries are null, which means deaths are not reported in the registry record. A null death entry should not be interpreted as a confirmed count of zero.

The study has a crossover design, but the adverse-event display does not explain how the at-risk counts relate to periods, how events were assigned to treatment, or how repeated exposure was handled. The counts should be read as posted, not added together or interpreted as unique participants across the trial. The record does not provide event definitions, severity grading beyond the serious category, or detail about duration of safety follow-up. These registry entries are limited safety reporting and do not establish the complete safety profile of either intervention.

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

The posted primary result is an estimated mean difference in postprandial triglyceride area under the curve between the sitagliptin and placebo phases. The mean was 27.1 mmol*h/L during sitagliptin and 30.0 mmol*h/L during placebo, with a reported difference of -2.9 mmol*h/L. The confidence interval from -9.262 to 3.462 mmol*h/L indicates substantial uncertainty around the direction and magnitude of that contrast. The unadjusted p-value was 0.372.

This result does not demonstrate that the treatment phases had identical outcomes. A confidence interval that includes zero is compatible with a range of values, including no difference. It also does not prove a difference in either direction. The result is best described as an estimate with uncertainty, rather than as a definitive positive or negative conclusion.

The endpoint is a summary of triglyceride exposure during the specified postprandial sampling period. It does not directly establish effects on fasting triglycerides, every lipoprotein measure, cardiovascular risk, or clinical events. The trial's background describes a potential rationale for studying postprandial lipemia, but the posted outcome is not a cardiovascular outcome. No secondary results are posted.

The registry describes men with type 2 diabetes, a per-protocol analysis, and a crossover comparison of the two treatment phases. The findings should be understood within that population, endpoint, and analysis description. The record does not give further participant characteristics, treatment sequence, washout details, or the precise statistical model. Those omissions constrain how broadly the result can be interpreted.

The serious adverse event entries show 0 events in each listed treatment category, with death counts not reported in the registry record. Such entries are limited to what the registry records and do not by themselves establish comparative safety. No medical decision should be based on this statistical summary alone.

16. Why This Trial Matters Statistically

This trial illustrates why a point estimate should not be interpreted without its uncertainty. The mean difference is -2.9 mmol*h/L, but the confidence interval extends from -9.262 to 3.462 mmol*h/L. That interval includes values on both sides of zero, so the estimate does not precisely resolve the direction of the mean contrast. The p-value of 0.372 adds a test-based summary but does not replace the effect estimate or interval.

The design also matters to statistical interpretation. A crossover study can compare measurements across treatment periods within participants, while the posted arm summaries report phase-specific means and standard deviations. The within-participant relationship is important to the precision of a paired comparison, yet the registry does not report it or the statistical model. This makes the posted confidence interval central to judging precision and prevents independent reconstruction from the marginal summaries alone.

The endpoint illustrates the difference between a biological measure and a clinical outcome. Postprandial triglyceride area under the curve summarizes measurements over a defined period. It can address the stated lipid question but does not directly answer whether cardiovascular outcomes differ. Keeping the endpoint and inference aligned avoids extending the result beyond what was measured.

Finally, the record shows how registry reporting shapes what can be assessed. It includes the primary outcome and serious adverse event counts, but does not report secondary results, detailed missing-data methods, or the trial's original sample-size rationale. These are not proof of shortcomings in the underlying conduct; they are limits on what can be evaluated from the registry record.

17. 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 registry record does not report the detailed statistical model, the test used for the primary comparison, or how the crossover pairing was incorporated. The treatment sequence, washout details, period effects, and carryover assessment are also not reported. These omissions matter because crossover estimates may depend on how period and within-participant structure are handled.

The primary outcome description identifies a per-protocol analysis, but does not define the analysis population in detail. The record does not report protocol deviations, participant flow, missing measurements, missing-data methods, or sensitivity analyses. Although the phase-specific results each list 36 observations, that does not resolve how incomplete outcome data or exclusions were managed.

The registry does not report the original sample-size justification, interim analyses, or a full statistical analysis plan. It lists one outcome measure and no secondary results or analysis records, so it is not possible to assess the completeness of outcome reporting or any multiplicity strategy from this record. The confidence interval method and confidence level are not specified in the registry record.

Generalizability is also bounded by the described population: men with type 2 diabetes. The record does not provide additional participant characteristics or details of the meal and sampling procedures. The endpoint concerns postprandial triglyceride exposure at the stated time points, not cardiovascular outcomes or all lipid measures. These limits should accompany interpretation of the numerical result.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effects of Sitagliptin on Postprandial Lipemia in Men With Type 2 Diabetes — NCT00660075.
• Registry record last updated 2012-12-11; status last verified 2012-11.
• Results first posted 2012-12-11.
• 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.

21. Record Summary

The completed Phase 3 trial was randomized, quadruple-masked, placebo-controlled, and crossover. It enrolled 36 men with type 2 diabetes and compared sitagliptin 100 mg/d for 6 weeks with placebo for 6 weeks. The primary outcome was postprandial plasma triglyceride area under the curve, measured at time 0, 2, 4, 6, and 8 hours at the end of both interventions.

The posted mean was 27.1 mmol*h/L for sitagliptin and 30.0 mmol*h/L for placebo. The mean difference was -2.9 mmol*h/L, with a confidence interval from -9.262 to 3.462 mmol*h/L and an unadjusted p-value of 0.372. The interval includes zero and indicates uncertainty about the direction and size of the contrast. Serious adverse event counts were 0 in each listed category; death counts are not reported in the registry record. No secondary results are posted.