Effects of Colesevelam on How the Body Responds to Insulin in Patients With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effects of Colesevelam on How the Body Responds to Insulin in Patients With Type 2 Diabetes (NCT00147745), a randomized phase 2 study. The primary outcome is Difference in Endogenous (Hepatic) Glucose Output During a High-dose Insulin Infusion From Baseline to After 12 Weeks of Treatment..

2. Clinical Question

The trial asked whether colesevelam changes insulin sensitivity in people with type 2 diabetes. Its primary measure was the change from baseline to 12 weeks in endogenous, or hepatic, glucose output during a high-dose insulin infusion. The registry describes a decrease in this measure after colesevelam treatment as indicative of greater liver sensitivity to insulin. This is a physiological outcome: it addresses a proposed mechanism rather than directly establishing a broader clinical effect.

The primary comparison was colesevelam 3.8g against matching placebo. An open-label insulin glargine group was also included. The registry lists change in oral glucose absorption and change in hemoglobin A1C as secondary outcomes. The posted primary result does not show a clear separation between colesevelam and placebo: the sponsor’s adjusted estimate has a confidence interval spanning values on either side of no difference, and its p-value is 0.550. The unadjusted recalculation likewise has an interval spanning no difference. These estimates are imprecise, so they do not establish that the groups were equivalent or that a meaningful difference was absent.

3. Trial Design

The registry describes a randomized, double-masked, parallel phase 2 study sponsored by Daiichi Sankyo. It lists 36 participants enrolled, a start of 2005-06, primary completion in 2008-01, and a completed status. The condition is type 2 diabetes. These details describe the registered design and timing; they do not establish the representativeness of participants or the conduct of the study beyond what is posted.

Three groups are described: colesevelam 3.8g daily for 12 weeks, matching placebo for 12 weeks, and open-label insulin glargine for 12 weeks. The presence of an open-label group is distinct from the registry’s overall double-masked designation. The registry does not give details of how masking was maintained for the other groups or explain the role of the insulin glargine group in the study’s primary comparison. Those design details are not reported in the registry record.

The primary endpoint was a continuous measure, with results expressed as least-squares means and standard errors. The posted arm counts for this endpoint are 15 for colesevelam, 13 for matching placebo, and 6 for open-label insulin glargine. The registry notes that one placebo participant did not have a valid post-randomization insulin clamp study. The analysis population is described as randomized participants who received medication and had a baseline and at least one post-randomization efficacy assessment, but the description ends before fully specifying the placebo group count. Further detail about exclusions, missing-data handling, and analysis-population definitions is not reported in the registry record.

The record lists five outcome measures and 13 analyses, while two secondary outcomes have posted numerical results. The original sample-size justification, intended power, planned effect size, and detailed statistical analysis plan are not reported in the registry record. The trial’s sample-size rationale therefore cannot be inferred from its enrollment or from the later planning illustrations.

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: Subjects were taken off any non-sulfonylurea oral anti-diabetic agents for a period of time appropriate to the drug. Qualified subjects were randomized to colesevelam hydrochloride, placebo, or open-label insulin glargine for 12 weeks. Difficulties enrolling into the insulin group caused the size of this group to decrease from 15 to 6 subjects.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The population analyzed was all randomized subjects who received medication and had a baseline and at least 1 post-randomization assessment of an efficacy variable. One placebo participant did not have a valid post-randomization insulin clamp study. Therefore, the number included in the placebo group for this analysis was 13.

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 Difference in Endogenous (Hepatic) Glucose Output During a High-dose Insulin Infusion From Baseline to After 12 Weeks of Treatment., measured over Baseline to 12 weeks.

The primary endpoint is the difference in endogenous, or hepatic, glucose output during a high-dose insulin infusion from baseline to after 12 weeks of treatment. It is measured in mg/kg/min. The registry reports least-squares mean changes with standard errors, rather than individual participant values. In the posted results, the colesevelam arm’s mean change is -0.06 and the matching-placebo arm’s is -0.011; the open-label insulin glargine arm’s is -0.01.

The rationale recorded for this measure is that a decrease in hepatic glucose output after colesevelam treatment is indicative of greater liver sensitivity to insulin. The infusion provides a specified physiological context for assessing glucose output, while the baseline-to-follow-up change focuses the comparison on change during treatment. This outcome is therefore intended to address a possible mechanism of action. It is not, by itself, a direct measure of every aspect of blood glucose control or of a patient’s overall clinical experience.

The registry does not provide further procedural detail about the clamp study or explain how the least-squares means were adjusted. Those details are not reported in the registry record. As with any outcome, interpretation depends on how the measure was collected, how missing observations were treated, and how the comparison was specified. The posted analysis and the stated endpoint definition support a focused reading: they concern measured hepatic glucose output during an insulin infusion over the stated treatment period.

7. Results Posted in the Registry

• Colesevelam 3.8g: mean -0.06 versus -0.01 on control; difference -0.05 (95% CI -0.19 to 0.09), standardized effect d = -0.26; z-test p = 0.484.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

The Acute Effect of Colesevelam (Multiple Doses) on Oral Glucose Absorption From Baseline to 12 Weeks

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Colesevelam 3.8g: mean -49.10 versus -4.70 on control; difference -44.40 (95% CI -146.51 to 57.71), standardized effect d = -0.32; z-test p = 0.394.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Colesevelam 3.8g and the control. The interval gives the range of effects that remain compatible with the data.

Change in Hemoglobin A1C Due to Effect of Colesevelam From Baseline to 12 Weeks

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Colesevelam 3.8g: mean -0.29 versus 0.16 on control; difference -0.45 (95% CI -1.17 to 0.27), standardized effect d = -0.46; z-test p = 0.219.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Colesevelam 3.8g and the control. The interval gives the range of effects that remain compatible with the data.

The two secondary outcomes extend the study’s mechanistic focus to oral glucose absorption and hemoglobin A1C. Across the posted sponsor comparisons, the estimates are uncertain: their confidence intervals include no difference. The overall pattern does not provide a clearly separated set of results that resolves the uncertainty in the primary comparison.

These outcomes are related to, but not the same as, hepatic glucose output during an insulin infusion. Oral glucose absorption concerns a different physiological process, while hemoglobin A1C is a measure of average glycemia over time. Considering several types of outcomes can provide a broader account of what was measured, but a secondary result does not automatically confirm or overturn the primary endpoint. In this record, the secondary analyses do not remove the uncertainty in the primary estimate.

There is also a general statistical reason to give unadjusted secondary findings less weight. When multiple outcomes and comparisons are examined, the chance of finding an apparently notable result by chance can increase. A prespecified hierarchy or multiplicity adjustment can help govern interpretation, but whether either was used is not reported in the registry record. The record lists five outcome measures and 13 analyses, while numerical results are posted for two secondary outcomes.

Secondary outcomes can be useful for generating hypotheses and describing the range of measured responses. They are less conclusive when their place in the testing plan, missing-data handling, and multiplicity strategy are unclear. The secondary findings should therefore be considered alongside the primary endpoint, the analysis methods, and the small group counts, rather than treated as independent confirmation of a treatment effect.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• The Acute Effect of Colesevelam (Multiple Doses) on Oral Glucose Absorption From Baseline to 12 Weeks
• Change in Hemoglobin A1C Due to Effect of Colesevelam From Baseline to 12 Weeks

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it records a quantity on a measurement scale rather than classifying each participant into a category. The registry reports least-squares mean changes and standard errors for each arm. A least-squares mean is a model-based group mean; the record identifies it as the reported parameter but does not describe the model used to produce each arm estimate. A standard error describes uncertainty in an estimated mean. It is not the same thing as the spread of individual participant measurements.

What does the arm comparison estimate?

The unadjusted recalculation contrasts the posted mean change for colesevelam with the posted mean change for matching placebo. Its mean difference is -0.049. The confidence interval is -0.186 to 0.088, and the unadjusted p-value is 0.484. The calculation is based on posted summary statistics, rather than participant-level data, and its pooled standard deviation is approximate. It is consequently a simple unadjusted comparison, not a replacement for the sponsor’s model-based analysis.

The registry reports ANCOVA results as mean differences in final values. ANCOVA is a regression approach that can compare outcome values while accounting for other variables included in the model, often including baseline outcome values. The registry identifies ANCOVA as the method, but the model covariates and further implementation details are not reported in the registry record. The sponsor’s reported primary comparison is 0.04, with a 95% confidence interval from -0.10 to 0.19 and p = 0.550. Since this is a reported final-value analysis and the recalculation uses posted changes, the estimates need not match exactly.

How should a confidence interval be read?

A confidence interval describes the uncertainty around an estimated difference under the assumptions of the analysis. It gives a range of values reasonably compatible with the data and the method; it is not a guarantee that the true effect lies within that particular interval. The primary recalculated interval includes zero, the value corresponding to no mean difference. The sponsor’s adjusted interval also includes zero. Both intervals extend to values on either side of zero, so the data do not precisely distinguish the direction of the comparative effect.

A p-value is calculated under a statistical model that assumes no difference for the comparison being tested. It describes how unusual data at least as discrepant as those observed would be under that assumption and the model. It does not quantify the probability that the null hypothesis is true, nor does it describe the size or practical importance of the estimate. The posted p-values should be interpreted alongside the estimates and intervals, rather than as stand-alone conclusions.

Why can the recalculation differ from the reported analysis?

The recalculation uses summary changes and an unadjusted comparison. The registry’s ANCOVA results concern mean differences in final values. Adjusting for baseline values and using a model can change an estimate and its uncertainty; analyses can also differ in their exact populations or assumptions. The registry does not provide enough model detail to reconstruct the ANCOVA. The posted recalculation is useful as a transparent descriptive check, but the registry’s own adjusted analysis is the relevant result for understanding the sponsor’s reported method.

The recalculated pooled standard deviation is estimated from the posted standard errors and group sizes, rather than calculated from participant-level observations. That approximation is one reason to distinguish the recalculation from a full analysis. The record’s stated primary analysis population also excludes participants without the described efficacy assessments, and one placebo participant lacked a valid post-randomization clamp study. Details of imputation or other missing-data procedures are not reported in the registry record.

What does randomization contribute?

Random allocation is intended to make treatment groups comparable on average at the time of assignment, including with respect to characteristics that may not have been measured. It does not ensure that every group will be balanced in a small trial, and it does not eliminate uncertainty from missing outcome data or measurement. The registry labels the study randomized, but the sequence-generation and allocation-concealment procedures are not reported in the registry record.

The double-masked designation applies alongside a separately described open-label insulin glargine arm. Masking can reduce the influence of expectations on care or outcome assessment, depending on how it is implemented. The record does not explain which participants or study staff were masked, or how masking related to the open-label arm. Those implementation details are not reported in the registry record.

Finally, analyses across several outcomes and comparisons raise the general issue of multiplicity: testing more hypotheses can increase the chance that at least one result appears notable by chance. The record lists five outcome measures and 13 analyses, but does not report whether a multiplicity adjustment or hierarchy was planned. General statistical practice is to interpret unadjusted secondary analyses cautiously, especially when a prespecified testing strategy is unavailable.

11. Confidence Intervals and Precision

Precision is reflected in the width of a confidence interval: a narrower interval indicates less statistical uncertainty under the analysis assumptions, while a wider interval leaves a broader range of compatible effects. For the primary unadjusted recalculation, the mean difference is -0.049 and the 95% confidence interval runs from -0.186 to 0.088. The sponsor’s ANCOVA estimate is 0.04, with a 95% confidence interval from -0.10 to 0.19. Each interval crosses zero, so neither analysis precisely establishes the direction of the difference.

The two intervals are not directly interchangeable. The recalculation compares arm-level changes, while the sponsor’s reported ANCOVA compares final values. The reported estimate and interval should be read as the sponsor’s analysis; the recalculation is an unadjusted estimate based on posted summary statistics. The registry does not provide enough detail about the ANCOVA model to reproduce it independently.

Limited group sizes are relevant to precision. The primary endpoint results include 15 participants in the colesevelam arm, 13 in the placebo arm, and 6 in the open-label insulin glargine arm. The registry also notes a placebo participant without a valid post-randomization clamp study. With limited observations, estimates may be sensitive to participant-level variation, and intervals can remain wide. The reported standard errors quantify uncertainty in the arm means, while confidence intervals for comparisons express uncertainty in the differences.

Precision is separate from whether a result meets a significance threshold. An interval spanning zero does not prove no effect, and the endpoints of an interval are not a range of effects equally likely under every model. The registry does not report a prespecified threshold for a clinically important difference, so statistical precision cannot be translated into a judgment about clinical importance from these results alone.

12. 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 original trial design. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each planning scenario uses the stated fraction of the observed effect as the effect to detect. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

For the primary comparison, planning for 100% of the observed effect gives 224 participants per arm for 80% power and 300 participants per arm for 90% power. Planning for 75% of the observed effect gives 398 participants per arm for 80% power and 533 participants per arm for 90% power. Planning for 50% of the observed effect gives 895 participants per arm for 80% power and 1,198 participants per arm for 90% power.

In plain language, the 80% power column represents the illustrative number per arm needed to have an 80% chance of detecting the specified effect under the planning assumptions if that effect were the true one. The 90% power column uses a higher target chance of detection under the same kind of assumptions. Detecting a smaller effect requires more participants, which is why the planning figures increase as the assumed fraction of the observed effect decreases.

These are planning illustrations, not observed power and not evidence that the trial was designed with these group sizes. They depend on the observed effect used as their starting point, which is uncertain in a small study. They should be read as conditional scenarios rather than as a recommendation or as a reconstruction of the trial’s original design.

13. 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 = -0.049; pooled standard deviation σ = 0.185.
2. n = 2 × (1.96 + 0.842)² × 0.185² / -0.049² = 223.7, rounded up to 224 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 <- -0.049                      # observed difference in means
sigma <- 0.184973594789177                     # 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(147745)
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 = 224 (80% power), 300 (90% power)
 75% of observed effect: n per arm = 398 (80% power), 533 (90% power)
 50% of observed effect: n per arm = 895 (80% power), 1,198 (90% power)
power.t.test: 224.7 
Simulated power at n = 224 per arm: 0.799 (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.

14. Analyses Reported by the Sponsor

The sponsor’s reported primary analysis is an ANCOVA comparison of final values for colesevelam and matching placebo. It gives a mean difference of 0.04, a 95% confidence interval from -0.10 to 0.19, and p = 0.550. The interval includes zero and spans values on either side of it, indicating uncertainty about the direction and size of the adjusted difference.

The unadjusted recalculation uses the posted arm-level changes instead. It gives a mean difference of -0.049, a 95% confidence interval from -0.186 to 0.088, and an unadjusted p-value of 0.484. These results are not identical because they concern different analysis approaches: an unadjusted comparison of changes versus the reported ANCOVA of final values. The registry does not give sufficient detail to reproduce the adjusted model or fully explain the relationship between its estimate and the change-based recalculation.

The recalculation is a check based on summary data, with an approximate pooled standard deviation. It should not be substituted for the sponsor’s analysis or described as a participant-level replication. Both approaches yield intervals that include no difference, but each must be interpreted according to its own estimand and assumptions. The sponsor’s analyses for the other reported comparisons are also recorded as ANCOVA results; those are distinct from unadjusted recalculations based on arm summaries.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, treatment comparisons, or statistical tests are considered. Even if each test is conducted appropriately on its own, examining many tests can increase the chance that at least one result looks unusual by chance. This is a general statistical concern and does not establish that any particular result in this trial is spurious.

The registry lists five outcome measures and 13 analyses. It identifies the primary endpoint and two secondary outcomes with posted results, but does not report a testing hierarchy, a multiplicity adjustment, or whether the analyses were prespecified in the statistical analysis plan. Those details are not reported in the registry record. The original sample-size justification, interim analyses, and complete statistical analysis plan are also not reported in the registry record. It is therefore not possible to determine from the record how the investigators intended to control the interpretation of multiple tests.

The primary result should be distinguished from secondary comparisons. A primary endpoint is generally the central outcome used to address the main trial question, whereas secondary outcomes add context or address additional questions. Without a reported hierarchy or adjustment strategy, unadjusted secondary p-values should be interpreted cautiously and should not be treated as if each were the sole planned test.

The registry describes an analysis population consisting of randomized participants who received medication and had baseline and at least one post-randomization efficacy assessment. It also states that one placebo participant did not have a valid post-randomization insulin clamp study. The endpoint counts are 15 for colesevelam, 13 for placebo, and 6 for insulin glargine. The record does not fully describe how the analysis population was derived from all randomized participants, how missing observations were handled, or whether sensitivity analyses were conducted. Those matters are not reported in the registry record. Missing-data assumptions can affect an estimate, particularly when outcome availability differs across participants, so the posted counts and stated population definition are important context for interpretation.

16. 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.

Most frequent non-serious adverse event terms

The registry reports zero serious adverse events in each listed group: the colesevelam group has 0 among 16 at risk, the matching-placebo group has 0 among 14 at risk, and the open-label insulin glargine group has 0 among 6 at risk. These are counts in the safety reporting groups, which are not identical to the primary endpoint analysis counts. The figures describe what was recorded in this trial and should not be taken as proof that serious events cannot occur.

The registry does not list serious adverse-event terms. The number of deaths is not reported in the registry record. For other adverse events, the record lists 14 affected participants among 16 at risk for colesevelam, 10 among 14 for matching placebo, and 4 among 6 for insulin glargine. Those counts are distinct from the serious adverse-event counts and do not provide detail about event type, severity, timing, or relatedness in the information posted.

Small groups and a limited observation period constrain what these figures can establish about safety. The reported period for each intervention is 12 weeks. The registry counts alone do not support broad comparisons of safety between treatments or conclusions about uncommon events.

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

The results describe measured changes in hepatic glucose output during a high-dose insulin infusion over 12 weeks, alongside two secondary outcomes. For the primary comparison, both the unadjusted recalculation and the sponsor’s reported ANCOVA have confidence intervals that include no difference. The estimates therefore leave uncertainty about the direction and size of the difference between colesevelam and matching placebo.

This does not demonstrate that the treatments are equivalent. Equivalence requires a defined margin and a design and analysis intended to test whether differences fall within that margin. The registry does not report such a margin or an equivalence objective. Nor does a p-value above a conventional significance threshold establish that there is no effect. It means that the reported test did not provide strong evidence against its no-difference assumption under the analysis used.

The primary endpoint is a physiological measure chosen to address insulin sensitivity in the liver. It is not the same as a broad assessment of clinical benefit, long-term outcomes, or every aspect of blood glucose control. The registry’s description says that a decrease in hepatic glucose output after colesevelam is indicative of greater liver sensitivity to insulin, but the posted results should be interpreted as estimates with uncertainty rather than as a complete account of clinical effects.

The trial enrolled 36 participants, with smaller numbers in the primary endpoint groups, and one placebo participant lacked a valid post-randomization clamp study. The posted estimates may therefore be imprecise. The record does not report the original sample-size rationale, the complete analysis plan, or the handling of missing data. These omissions limit how confidently the results can be connected to broader populations or used to answer questions beyond the registered measurements and follow-up period.

Accordingly, the statistical meaning is narrow: the registry reports no clearly separated primary comparison, and its interval estimates permit differences in either direction. That is a statement about the evidence and its precision, not a recommendation about treatment or a conclusion about clinical use.

18. Why This Trial Matters Statistically

This trial illustrates why physiological endpoints, analysis methods, and uncertainty need to be considered together. The primary measure is specific: hepatic glucose output during a high-dose insulin infusion. The registry links a decrease in that measure to greater liver sensitivity to insulin, making the endpoint relevant to a proposed mechanism. But a mechanistic measure does not automatically answer broader questions about blood glucose control or clinical outcomes.

The comparison also shows why an adjusted sponsor analysis and a summary-statistic recalculation should be kept distinct. The sponsor reports an ANCOVA comparison of final values, while the unadjusted recalculation contrasts posted changes. They answer related but not identical statistical questions. Reporting both estimates can aid transparency, provided their different bases are clear and neither is mistaken for a direct replication of the other.

The confidence intervals matter as much as the point estimates. The primary intervals span no difference and values in either direction, so the data do not precisely establish which direction the comparison favors. A non-significant p-value is not evidence of equivalence, and a point estimate alone does not show the range of effects compatible with the data.

The record also makes visible limits common to small trials: modest endpoint group counts, a missing valid clamp study, multiple analyses, and incomplete reporting of the original statistical plan. These features do not invalidate the results, but they shape the strength and scope of conclusions that can be drawn. The planning illustrations further show how assumptions about the effect to detect influence the number of participants required, while not revealing the trial’s original design rationale.

19. 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.

Several trial-specific limitations affect interpretation. The primary endpoint analysis includes 15 participants in the colesevelam group, 13 in the matching-placebo group, and 6 in the open-label insulin glargine group. The registry reports that one placebo participant lacked a valid post-randomization insulin clamp study. These counts limit precision, especially for comparisons involving the smaller insulin glargine group.

The primary outcome is a physiological measure during an insulin infusion. It addresses the stated mechanism but does not directly cover all aspects of glucose control or clinical outcomes. The registry lists changes in oral glucose absorption and hemoglobin A1C as secondary outcomes, but only two secondary outcomes have posted numerical results. The record does not report how the outcomes were prioritized beyond identifying the primary endpoint.

Important methodological details are not reported in the registry record: the original sample-size justification, detailed model specification, missing-data methods, randomization procedures, allocation concealment, and the implementation of masking. The record also does not report whether interim analyses or multiplicity procedures were planned. Without those details, the reported estimates cannot be fully reconstructed or evaluated against a complete prespecified analysis plan.

The sponsor’s ANCOVA results and the unadjusted recalculations use different approaches and should not be treated as identical estimates. The recalculation relies on posted summary statistics and an approximate pooled standard deviation. The registry does not provide participant-level data or enough model detail to reproduce the sponsor’s adjusted analysis. Finally, the intervention period is 12 weeks; longer-term effects and outcomes are not addressed by the posted results.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Effects of Colesevelam on How the Body Responds to Insulin in Patients With Type 2 Diabetes — NCT00147745.
• Registry record last updated 2014-04-02; status last verified 2014-03.
• Results first posted 2009-12-04.
• 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.

23. Record Summary

The registry describes a completed randomized, double-masked, parallel phase 2 study sponsored by Daiichi Sankyo, with 36 participants enrolled and people with type 2 diabetes as the study population. The registered interventions were colesevelam 3.8g daily for 12 weeks, matching placebo for 12 weeks, and open-label insulin glargine for 12 weeks.

The primary endpoint was change from baseline to 12 weeks in hepatic glucose output during a high-dose insulin infusion. The sponsor’s ANCOVA comparison of colesevelam with placebo has a 95% confidence interval spanning no difference. An unadjusted recalculation based on posted changes also has an interval spanning no difference. Both should be read as uncertain estimates, and the two analyses are not identical in method.

The registry posts two secondary outcomes and reports zero serious adverse events in each listed group. The original sample-size justification, detailed statistical plan, and missing-data methods are not reported in the registry record.