Effect of Fermentable Carbohydrate on Glucose Homeostasis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Fermentable Carbohydrate on Glucose Homeostasis (NCT01841073), a randomized na study. The primary outcome is Glycaemic Control.

2. Clinical Question

The registry frames the study around whether fermentable carbohydrate might influence appetite, food intake, weight, and blood glucose control in people at risk of type 2 diabetes. Inulin is the fermentable carbohydrate intervention; cellulose is the comparison intervention. The posted primary outcome is glycaemic control at 9 weeks. It is not a direct measure of diabetes prevention, and the record does not report a diabetes-incidence outcome.

The statistical comparison is between the recorded outcome means for the inulin and cellulose groups. The reported arm means are -0.23 for inulin and 0.44 for cellulose, in mmol//l. The registry also reports a sponsor ANCOVA comparison, but does not give its adjusted effect estimate or confidence interval. The clinical question should therefore be read as a comparison of this trial’s recorded nine-week measure, rather than as a broader conclusion about long-term health outcomes or prevention.

3. Trial Design

The registry describes a randomized, triple-masked, parallel-group trial. Participants were allocated to inulin or cellulose, and the groups were followed in parallel. The trial is marked completed, with enrollment recorded as 44. The outcome record gives an analysis count of 20 for inulin and 19 for cellulose. The registry does not explain the relationship between the enrollment figure and the outcome counts, or report why the counts differ.

The inulin regimen increased from 10g for 2 weeks to 20g for the next 2 weeks, then to 30g for the remainder of the investigations. The cellulose regimen followed the same dose schedule: 10g a day for 2 weeks, 20g a day for 2 weeks, and 30g a day for the rest of the study. The record does not describe other components of the study procedures or specify what triple masking meant in practice.

The primary outcome is labelled glycaemic control and has a 9-week timeframe. An oral glucose tolerance test using 75 g of glucose was performed to clarify glycaemic status. The record lists a continuous outcome summarized as a mean with standard error. It also lists food intake in grams at 9 weeks as a secondary outcome.

The record gives the phase as NA, the allocation as randomized, and the model as parallel. It does not report the trial’s original sample-size justification, the intended analysis population, allocation concealment procedures, or a detailed protocol for handling departures from the assigned regimen. The record identifies Imperial College London as sponsor. The study start is listed as 2011-03 and primary completion as 2013-08.

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 Glycaemic Control, measured over 9 weeks.

The primary outcome is titled glycaemic control and is measured at 9 weeks. The registry says an oral glucose tolerance test using 75 g of glucose was performed to clarify glycaemic status. In general, an oral glucose tolerance test assesses the body’s handling of a glucose challenge. It can provide a measure related to glucose regulation at the time of testing, which is relevant to a study in people with prediabetes.

The posted outcome is continuous and summarized as a mean with standard error. The recorded mean is -0.23 for the inulin group and 0.44 for the cellulose group, in mmol//l. The registry does not explain what change or transformation, if any, these values represent, and does not describe a baseline value for this outcome. Accordingly, the values should be interpreted as the posted outcome summaries rather than assigned a more specific meaning than the record gives them.

An oral glucose tolerance test is relevant to the stated question because the study concerns glycaemic control and the registry explicitly uses the test to clarify glycaemic status. It is not, by itself, a measure of long-term diabetes incidence, weight management, or all aspects of glucose homeostasis. The registry record does not report those as the primary endpoint at 9 weeks.

7. Results Posted in the Registry

• Cellulose: mean 0.44 versus -0.23 on control; difference 0.67 (95% CI 0.09 to 1.25), standardized effect d = 0.74; z-test p = 0.023.

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.

Food Intake in Gram at 9 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:
  • Cellulose: mean -0.47 versus -127.00 on control; difference 126.53 (95% CI 27.22 to 225.84), standardized effect d = 0.79; z-test p = 0.013.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Cellulose than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry’s secondary outcome is food intake in grams at 9 weeks. Across the posted primary and secondary comparisons, the numerical contrasts point in the same direction when expressed as cellulose relative to inulin. This broad consistency is relevant as a pattern across outcomes, but it does not establish that the outcomes measure the same process or that one result explains the other.

Food intake is related to the study’s stated interest in appetite and intake, while the primary outcome concerns glycaemic control measured using an oral glucose tolerance test. A difference in food intake could be relevant to the broader rationale, but the registry does not report an analysis establishing that food intake mediated or caused the primary outcome difference. Nor does it report a joint analysis of these outcomes.

The sponsor reports ANCOVA analyses for the primary and secondary comparisons, and the record also includes unadjusted recalculations. Secondary findings generally carry less weight than the primary result when multiple outcomes are examined, particularly when the registry does not state whether multiplicity was addressed or identify a prespecified hierarchy. Repeated testing can increase the chance of at least one apparently notable result arising by chance. The record does not report an adjustment for multiple comparisons.

The registry lists three outcome measures, while one secondary result is posted. It does not explain the status or results of other outcome measures. The posted secondary finding should therefore be viewed as supporting context rather than independent confirmation of the primary result. It does not extend the primary result to weight management, long-term diabetes prevention, or outcomes not reported in the registry.

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

• Food Intake in Gram at 9 Weeks

9. Statistical Interpretation

10. Statistical Methods Explained

The trial is described as randomized, triple-masked, and parallel. Random allocation is intended, in general, to make treatment groups comparable on average at the point of assignment, including for characteristics that may not be measured. Masking is intended to reduce the influence of expectations on care, measurement, or reporting. The record does not describe who was masked or how allocation was concealed, so those design features cannot be evaluated in detail from the registry entry.

What does a mean and standard error describe?

The primary outcome is continuous, and each arm is summarized by a mean and standard error. In general, a mean represents the average of the observed values in a group. The standard error describes the statistical uncertainty in estimating a group mean; it is not the same as the spread of individual participant values. The registry labels the arm-level dispersion as standard error and gives 0.17 for inulin and 0.24 for cellulose. The posted group sizes are 20 and 19, respectively.

The recalculated comparison is a simple unadjusted mean difference based on the posted summaries. Its reported direction is cellulose minus inulin: the estimated difference is 0.67. The interval is 0.094 to 1.246, and the unadjusted p-value is 0.023. This calculation does not account for covariates or other features of an adjusted model. It is a way to describe the observed summary-statistic contrast, not a substitute for the sponsor’s analysis.

What does ANCOVA contribute?

The sponsor reports an ANCOVA analysis. In general, analysis of covariance compares groups while adjusting for one or more covariates included in the model. Depending on the covariates and model specification, adjustment can change the estimated group contrast and its uncertainty. The registry reports p = 0.005 for the superiority analysis, but does not give the adjusted effect estimate, confidence interval, covariates, model diagnostics, or exact analysis population. It is therefore not possible to reconstruct or fully assess that analysis from the registry record.

The record’s comparison is classified as superiority. In general, a superiority analysis tests whether one group differs from another in a specified direction or whether the groups differ under a prespecified hypothesis. The registry does not report the detailed hypothesis, the significance threshold used in the trial’s own plan, or whether any adjustments were made for multiple outcomes. The sample-size planning illustration uses a two-sided significance level of 0.05, but those assumptions are explicitly for the illustration and are not reported as the trial’s original design assumptions.

How should the confidence interval and p-value be read?

A confidence interval expresses uncertainty around an estimated effect under a statistical procedure. In general, narrower intervals indicate greater precision, while wider intervals are compatible with a broader range of values. Here, the interval for the unadjusted primary contrast extends from 0.094 to 1.246. The confidence level is not reported in the registry record. The interval should be considered alongside the point estimate, and it does not indicate that every value inside it is equally plausible.

The unadjusted p-value of 0.023 and the sponsor ANCOVA p-value of 0.005 arise from different analysis descriptions. A p-value does not quantify clinical importance. Nor does it convey the probability that a result will replicate. The effect estimate, uncertainty, design, outcome definition, and possible multiplicity all matter when interpreting the evidence.

Why can over-dispersion matter?

In general, over-dispersion means observed variability is greater than a model expects. If a model assumes less variability than is present, standard errors can be too small and confidence intervals can be too narrow, making statistical evidence look stronger than warranted. This trial’s primary outcome is continuous and the posted comparison is a mean difference; the registry does not report an over-dispersion assessment or identify over-dispersion as a concern. The concept is relevant as a general modeling consideration, not as a specific finding about this trial.

Overall, the reported ANCOVA is the sponsor’s inferential analysis, while the unadjusted recalculation is an uncomplicated comparison of the posted means. The registry does not provide the protocol or sufficient modeling detail to judge every assumption. The two results should be distinguished by method rather than treated as interchangeable estimates.

11. Confidence Intervals and Precision

The unadjusted primary contrast is 0.67, with a confidence interval from 0.094 to 1.246. This interval communicates uncertainty about the mean difference under the calculation used for the posted summary statistics. Its full range is on the positive side of the cellulose-minus-inulin contrast, but it still includes a range of possible effect sizes. The confidence level is not reported in the registry record, so the interval should not be assigned a particular coverage level.

Precision is distinct from statistical significance. In general, an estimate can have a p-value below a conventional threshold while still being imprecise enough to leave meaningful uncertainty about the magnitude. The interval is therefore more informative about the range of values compatible with the recalculation than the p-value alone. The trial’s outcome counts are 20 for inulin and 19 for cellulose, and the registry does not explain missingness or exclusions that could affect precision or representativeness.

The sponsor ANCOVA result has p = 0.005, but its adjusted estimate and interval are not reported. This prevents a direct assessment of the precision of the sponsor’s adjusted effect. The recalculated interval should not be presented as though it were the interval for the ANCOVA result. The registry record also does not provide a confidence level for the interval reported with the unadjusted comparison.

Even where an interval excludes zero, it does not establish clinical importance, persistence, or generalizability. Those judgments require a clear outcome scale and context, and the registry’s description does not specify the interpretation of the reported values beyond the glycaemic-control label and OGTT description. The interval expresses uncertainty in the estimate; it does not resolve those broader questions.

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 illustrations give participants per arm, not total events. They are based on a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is expressed as a stated fraction of the observed effect. The registry does not report whether these assumptions were used to design the original trial; the original sample-size justification and design assumptions are not reported in the registry record.

For the primary outcome, the displayed planning values are 29 participants per arm for 80% power and 39 participants per arm for 90% power when planning for 100% of the observed effect. When planning for 75% of the observed effect, the corresponding figures are 52 and 70 participants per arm. For 50% of the observed effect, they are 116 and 156 participants per arm. These are illustrations based on the observed estimate, not a reconstruction of the trial’s original target.

In plain terms, 80% power means that, under the stated assumptions and if the specified effect were the true effect, the design would have an 80% chance of detecting it using the chosen statistical test. The 90% column represents a higher chance of detecting that effect under those same assumptions, and generally requires more participants. Planning for a smaller effect also requires more participants, because smaller differences are harder to distinguish from random variation.

These figures are not observed power. They are planning illustrations, and they should not be read as evidence that the completed trial had a particular probability of finding its result. The curve and the displayed estimates describe how hypothetical sample requirements vary with the target effect and power, not what happened to participants in the trial.

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 = 0.670; pooled standard deviation σ = 0.911.
2. n = 2 × (1.96 + 0.842)² × 0.911² / 0.670² = 29.0, rounded up to 29 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 <- 0.67                      # observed difference in means
sigma <- 0.910616067078559                     # 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(1841073)
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 = 29 (80% power), 39 (90% power)
 75% of observed effect: n per arm = 52 (80% power), 70 (90% power)
 50% of observed effect: n per arm = 116 (80% power), 156 (90% power)
power.t.test: 30 
Simulated power at n = 29 per arm: 0.791 (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 own reported analysis is an ANCOVA superiority analysis with p = 0.005. The registry does not report its adjusted estimate, confidence interval, covariates, or other model details. The unadjusted recalculation instead compares the posted group means directly and gives a mean difference of 0.67, with a confidence interval from 0.094 to 1.246 and an unadjusted p-value of 0.023.

These results answer related but not identical statistical questions. ANCOVA can adjust for covariates, while the recalculation uses only the posted summaries. The registry indicates no discrepancy between the sponsor result and the recalculation. The sponsor’s analysis therefore remains the primary reported analysis; the recalculation offers a transparent, unadjusted account of the arm summaries and should not be treated as a replacement for the sponsor model.

Because the registry does not provide the ANCOVA estimate or its interval, the size and precision of the sponsor-adjusted contrast cannot be compared directly with the unadjusted estimate. The p-values alone do not reveal how much adjustment changed the estimated effect, or whether differences arose from covariate adjustment or other modeling choices. Further explanation is not reported in the registry record.

15. Multiplicity, Missing Data and Analysis Populations

The registry reports a primary outcome and a secondary outcome, and lists three outcome measures. It reports an ANCOVA analysis for the primary and posted secondary outcome. The record does not state whether these outcomes were prespecified in a statistical analysis plan, whether the study used a hierarchy of hypotheses, or whether p-values were adjusted for multiple testing. It also does not explain the status of the outcome measures for which results are not posted.

In general, when multiple outcomes or analyses are tested, the chance of obtaining at least one apparently strong result by chance can increase. Multiplicity procedures, such as prespecified testing sequences or statistical adjustments, can address that issue. The record does not report that any such approach was used. The secondary result should consequently be interpreted with more caution than a single prespecified primary test, and the record does not support treating the set of results as a multiplicity-controlled family.

Missing-data handling is not reported in the registry record. The posted primary outcome counts are 20 for inulin and 19 for cellulose, compared with enrollment of 44. The registry does not say whether the difference reflects missing outcomes, exclusions, withdrawals, or another reason. It does not report imputation, sensitivity analyses, or the number randomized to each group in a way that clarifies the analysis population.

The analysis population is also not described. The record does not state whether analyses followed intention-to-treat, per-protocol, or another rule, and does not report how participants who did not follow the assigned regimen were handled. The registry gives no interim-analysis information, stopping rule, or detailed statistical analysis plan; these details are not reported in the registry record. Their absence limits assessment of the prespecification and robustness of the reported inferences, but does not establish that any particular procedure was or was not used.

16. 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.

For inulin, the registry reports 0 serious adverse events among 20 participants at risk and 0 other adverse events among 20 participants at risk; deaths are recorded as 0. For cellulose, it reports 0 serious adverse events among 19 participants at risk and 0 other adverse events among 19 participants at risk; deaths are recorded as 0. No serious adverse event terms are listed.

These entries describe the adverse-event counts posted in the registry. They do not establish that no symptoms or less formally recorded events occurred, and the record does not provide event definitions, monitoring details, or a narrative of how safety information was collected. The at-risk counts also do not explain the relationship between enrollment and the outcome analysis counts. The figures should therefore be read as the registry’s recorded counts, not as a complete characterization of all possible harms or of longer-term safety.

The safety record does not provide a basis for comparing uncommon events, and zero recorded events in these groups cannot rule out events that were not observed or reported. No medical conclusion follows from these counts alone.

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

The registry records a randomized comparison of inulin and cellulose in people with prediabetes, with glycaemic control measured at 9 weeks using an oral glucose tolerance test with 75 g of glucose. The posted means differ, and the sponsor reports an ANCOVA superiority analysis with p = 0.005. The unadjusted recalculation also gives a positive cellulose-minus-inulin contrast, with an interval reported from 0.094 to 1.246.

These findings describe the recorded outcome in this trial. They do not establish that inulin prevents type 2 diabetes, changes long-term risk, or produces sustained differences. Diabetes incidence is not reported as an outcome, and the primary assessment is at 9 weeks. The registry also does not establish that the measured contrast is clinically important, because it does not describe the outcome scale sufficiently to define the practical meaning of the values.

The comparison is between two dietary supplements with dose schedules that rise over the study period. It does not compare inulin with no supplement, and the registry does not report results for every outcome described in its background rationale. The primary outcome should not be generalized to weight management, appetite, or other effects unless those outcomes are directly reported and appropriately analyzed.

The reported p-values do not show the probability that the finding is true or that it will replicate. The interval around the unadjusted estimate reflects uncertainty under that calculation, while the sponsor’s adjusted estimate and interval are not reported. The registry also does not report missing-data procedures, analysis populations, or multiplicity handling. These limits matter when judging how robustly the result answers the question.

Accordingly, the record supports describing a statistical difference in the posted nine-week glycaemic-control outcome under the reported analyses. It does not, on its own, establish clinical benefit, harm, causation through a proposed biological mechanism, or effectiveness in populations beyond those studied. Those broader conclusions are not reported in the registry record.

18. Why This Trial Matters Statistically

This trial is statistically informative because it pairs a randomized comparison with a continuous clinical outcome and reports both a sponsor ANCOVA p-value and arm-level summaries that permit an unadjusted comparison. The distinction between the sponsor’s adjusted model and a summary-statistic recalculation illustrates why an analysis result should be read together with its method. They are related but not interchangeable quantities.

The reported group sizes are modest, and the interval for the unadjusted contrast communicates uncertainty about the magnitude even though the interval lies on one side of zero. This is a useful reminder that statistical evidence and precision are related but different: a p-value gives one view of compatibility with a null model, while an interval displays a range of effect estimates associated with the calculation.

The record also illustrates the importance of transparent reporting. The sponsor’s adjusted estimate and interval, original sample-size justification, missing-data approach, analysis population, and multiplicity procedures are not reported in the registry record. Without those details, readers cannot fully assess the analysis plan or reproduce the adjusted comparison. The planning illustrations are helpful for understanding how sample requirements depend on target effect and power, but they are not the trial’s original design rationale.

Finally, the primary outcome is a nine-week measure, while the study background refers to broader goals such as appetite, weight management, and diabetes prevention. Statistical interpretation needs to remain aligned with what was measured and reported. The trial’s record can inform understanding of a specific short-term comparison without resolving longer-term or wider clinical questions.

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 limitations are specific to the information recorded for this trial. The registry gives the primary outcome title, unit, timeframe, and OGTT description, but does not explain the scale or transformation underlying the negative and positive mean values. This makes it difficult to translate the estimated contrast into a clinical interpretation beyond the recorded group comparison.

The registry reports the sponsor’s ANCOVA p-value but not the adjusted estimate, interval, covariates, or model diagnostics. The unadjusted recalculation is therefore the only posted effect estimate with an interval, and it should not be mistaken for the adjusted sponsor result. Details of allocation concealment and the practical implementation of triple masking are also not reported.

The outcome counts are 20 and 19, while enrollment is 44. The record does not explain the difference, report participant flow, identify withdrawals, or describe missing-data handling. It also does not identify the analysis population or how adherence to the assigned dose schedule was handled. These omissions prevent assessment of possible attrition or exclusion effects.

The registry does not report the original sample-size justification or trial-specific assumptions, interim analyses, a full statistical analysis plan, or multiplicity procedures. It lists three outcome measures but posts one secondary result, without explaining the status of the other measures. These gaps constrain assessment of prespecification and selective reporting. The primary endpoint is measured at 9 weeks; longer-term outcomes, including diabetes incidence, are not reported in the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Effect of Fermentable Carbohydrate on Glucose Homeostasis — NCT01841073.
• Registry record last updated 2019-11-25; status last verified 2019-11.
• Results first posted 2019-11-01.
• 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

NCT01841073, titled “Effect of Fermentable Carbohydrate on Glucose Homeostasis,” is recorded as a completed randomized, triple-masked, parallel trial sponsored by Imperial College London. The registry lists enrollment as 44 and the condition as prediabetes.

The trial compared increasing-dose schedules of inulin and cellulose. The primary outcome was glycaemic control at 9 weeks, assessed using an oral glucose tolerance test with 75 g of glucose. The sponsor reports an ANCOVA superiority result with p = 0.005. The unadjusted recalculation reports a mean difference of 0.67 for cellulose compared with inulin, with a confidence interval from 0.094 to 1.246 and p = 0.023.

Food intake in grams at 9 weeks is the posted secondary outcome. The registry does not report the original sample-size justification, missing-data procedures, analysis population, multiplicity handling, or the adjusted ANCOVA estimate and interval.